From adrian@olddog.co.uk Mon Jun 17 10:51:46 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 934F721F9D7C for ; Mon, 17 Jun 2013 10:51:46 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.583 X-Spam-Level: X-Spam-Status: No, score=-2.583 tagged_above=-999 required=5 tests=[AWL=0.016, BAYES_00=-2.599] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 0X5CI-ZA4nO8 for ; Mon, 17 Jun 2013 10:51:40 -0700 (PDT) Received: from asmtp4.iomartmail.com (asmtp4.iomartmail.com [62.128.201.175]) by ietfa.amsl.com (Postfix) with ESMTP id 7F4F021F9D63 for ; Mon, 17 Jun 2013 10:51:37 -0700 (PDT) Received: from asmtp4.iomartmail.com (localhost.localdomain [127.0.0.1]) by asmtp4.iomartmail.com (8.13.8/8.13.8) with ESMTP id r5HHpaEX017240 for ; Mon, 17 Jun 2013 18:51:36 +0100 Received: from 950129200 (dsl-sp-81-140-15-32.in-addr.broadbandscope.com [81.140.15.32]) (authenticated bits=0) by asmtp4.iomartmail.com (8.13.8/8.13.8) with ESMTP id r5HHpWav017199 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO) for ; Mon, 17 Jun 2013 18:51:35 +0100 From: "Adrian Farrel" To: Date: Mon, 17 Jun 2013 18:51:32 +0100 Message-ID: <02aa01ce6b83$4fd21960$ef764c20$@olddog.co.uk> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Mailer: Microsoft Outlook 14.0 Thread-Index: Ac5rgxBKSYo0jotRQq28AmcXytnEVw== Content-Language: en-gb Subject: [nsc] Opening message X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list Reply-To: adrian@olddog.co.uk List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 17 Jun 2013 17:51:46 -0000 Testing the archives From mohamed.boucadair@orange.com Tue Jun 18 01:42:58 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C80AF21F9C8C for ; Tue, 18 Jun 2013 01:42:58 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.193 X-Spam-Level: X-Spam-Status: No, score=-2.193 tagged_above=-999 required=5 tests=[AWL=0.055, BAYES_00=-2.599, HELO_EQ_FR=0.35, UNPARSEABLE_RELAY=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id MfZgWxYeU-QG for ; Tue, 18 Jun 2013 01:42:54 -0700 (PDT) Received: from relais-inet.francetelecom.com (relais-ias91.francetelecom.com [193.251.215.91]) by ietfa.amsl.com (Postfix) with ESMTP id 7603B21F9C87 for ; Tue, 18 Jun 2013 01:42:54 -0700 (PDT) Received: from omfedm07.si.francetelecom.fr (unknown [xx.xx.xx.3]) by omfedm11.si.francetelecom.fr (ESMTP service) with ESMTP id 17EFA3B4DB2 for ; Tue, 18 Jun 2013 10:42:53 +0200 (CEST) Received: from PUEXCH51.nanterre.francetelecom.fr (unknown [10.101.44.31]) by omfedm07.si.francetelecom.fr (ESMTP service) with ESMTP id F2EC74C076 for ; Tue, 18 Jun 2013 10:42:52 +0200 (CEST) Received: from PUEXCB1B.nanterre.francetelecom.fr ([10.101.44.13]) by PUEXCH51.nanterre.francetelecom.fr ([10.101.44.31]) with mapi; Tue, 18 Jun 2013 10:42:52 +0200 From: To: "nsc@ietf.org" Date: Tue, 18 Jun 2013 10:42:51 +0200 Thread-Topic: I-D Action: draft-boucadair-network-function-chaining-00.txt Thread-Index: Ac5rZQg6UnH78S3KSNeD/gvMOIgmWAAmnOAA Message-ID: <94C682931C08B048B7A8645303FDC9F36ED99E6197@PUEXCB1B.nanterre.francetelecom.fr> References: <20130617141430.21635.51377.idtracker@ietfa.amsl.com> In-Reply-To: <20130617141430.21635.51377.idtracker@ietfa.amsl.com> Accept-Language: fr-FR Content-Language: fr-FR X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: fr-FR Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-PMX-Version: 5.6.1.2065439, Antispam-Engine: 2.7.2.376379, Antispam-Data: 2013.5.21.113319 Cc: JACQUENET Christian OLNC/OLN Subject: [nsc] TR: I-D Action: draft-boucadair-network-function-chaining-00.txt X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 18 Jun 2013 08:42:59 -0000 Dear all, We have submitted this new I-D. Comments are more than welcome. Cheers, Med -----Message d'origine----- De=A0: i-d-announce-bounces@ietf.org [mailto:i-d-announce-bounces@ietf.org]= De la part de internet-drafts@ietf.org Envoy=E9=A0: lundi 17 juin 2013 16:15 =C0=A0: i-d-announce@ietf.org Objet=A0: I-D Action: draft-boucadair-network-function-chaining-00.txt A New Internet-Draft is available from the on-line Internet-Drafts director= ies. Title : Differentiated Network-Located Function Chaining Framewo= rk Author(s) : Mohamed Boucadair Christian Jacquenet Filename : draft-boucadair-network-function-chaining-00.txt Pages : 16 Date : 2013-06-17 Abstract: IP networks rely more and more on the combination of advanced functions (besides the basic routing and forwarding functions). This document defines a solution to enforce Network-Located Function Chaining (NLFC) with minimum requirements on the underlying network. The proposed solution allows for Differentiated Forwarding (DiffForward): packets are classified at the entry point of an NLFC- enabled network, and are then forwarded on a per NLFC map basis. The IETF datatracker status page for this draft is: https://datatracker.ietf.org/doc/draft-boucadair-network-function-chaining There's also a htmlized version available at: http://tools.ietf.org/html/draft-boucadair-network-function-chaining-00 Internet-Drafts are also available by anonymous FTP at: ftp://ftp.ietf.org/internet-drafts/ _______________________________________________ I-D-Announce mailing list I-D-Announce@ietf.org https://www.ietf.org/mailman/listinfo/i-d-announce Internet-Draft directories: http://www.ietf.org/shadow.html or ftp://ftp.ietf.org/ietf/1shadow-sites.txt From paulq@cisco.com Tue Jun 18 10:27:45 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D4D7311E80E6 for ; Tue, 18 Jun 2013 10:27:45 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.299 X-Spam-Level: X-Spam-Status: No, score=-10.299 tagged_above=-999 required=5 tests=[AWL=0.300, BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id WUjpHavZVuaX for ; Tue, 18 Jun 2013 10:27:41 -0700 (PDT) Received: from mtv-iport-2.cisco.com (mtv-iport-2.cisco.com [173.36.130.13]) by ietfa.amsl.com (Postfix) with ESMTP id 72D2311E80F1 for ; Tue, 18 Jun 2013 10:27:41 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=729; q=dns/txt; s=iport; t=1371576461; x=1372786061; h=from:content-transfer-encoding:subject:message-id:date: to:mime-version; bh=ZfZvlWnalRK0knN4GdlI4cDsXHC/UMnbU3SE3GGGKSw=; b=EOwbeQhROSszkF+ktB/HZuDVWUsi6bEdAb+oADXWT8dXPEKRO4F33Ojs 3MsP8qQZC3VePG06fRZR15VKjZa7cacYLDlPOzWHW9ofNZzE+quWTYexV jgkslD4V+BggkV2xpXS1F1Z/k+Fx2gQAHEK/J+SlLTczGQN26om4rQPSt M=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AmEGAAaUwFGrRDoJ/2dsb2JhbABZgwkxAYMCr22NahZ0gxiJaZpJoEqPWIJqYQOJII4hgSmEd4QOhxWDLxw X-IronPort-AV: E=Sophos;i="4.87,890,1363132800"; d="scan'208";a="83817541" Received: from mtv-core-4.cisco.com ([171.68.58.9]) by mtv-iport-2.cisco.com with ESMTP; 18 Jun 2013 17:27:37 +0000 Received: from [10.155.64.54] ([10.155.64.54]) by mtv-core-4.cisco.com (8.14.5/8.14.5) with ESMTP id r5IHRa2F026735 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO) for ; Tue, 18 Jun 2013 17:27:36 GMT From: Paul Quinn Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: quoted-printable Message-Id: Date: Tue, 18 Jun 2013 10:27:36 -0700 To: nsc@ietf.org Mime-Version: 1.0 (Mac OS X Mail 6.5 \(1508\)) X-Mailer: Apple Mail (2.1508) Subject: [nsc] Network Service Chaining: Welcome and Let's Get Started X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 18 Jun 2013 17:27:46 -0000 Hello, Thank you for joining the Network Service Chaining (NSC) mailing list. = It=92s great to see so much interest in the topic! A BoF has been requested for IETF 87 with the intention of forming a = dedicated working group that will focus on all aspects of service chain = creation and associated protocol definitions.=20 A detailed BoF description as well as a proposed working group charter = is posted on the IETF BoF page (http://tools.ietf.org/bof/trac/).=20 In addition to the BoF request, several relevant drafts have been posted = and we invite you to please review them and send any = comments/suggestions. We look forward to lively and productive discussions on this list :) Thanks, Paul= From ietf-secretariat@ietf.org Tue Jun 18 14:59:39 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9DC6E21E80BC; Tue, 18 Jun 2013 14:59:39 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -102.6 X-Spam-Level: X-Spam-Status: No, score=-102.6 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, NO_RELAYS=-0.001, USER_IN_WHITELIST=-100] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id yBB3MfzfYMKj; Tue, 18 Jun 2013 14:59:38 -0700 (PDT) Received: from ietfa.amsl.com (localhost [IPv6:::1]) by ietfa.amsl.com (Postfix) with ESMTP id E52F421E80B7; Tue, 18 Jun 2013 14:59:38 -0700 (PDT) MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable From: IETF Secretariat To: IETF Announcement List X-Test-IDTracker: no X-IETF-IDTracker: 4.51.p2 Message-ID: <20130618215938.28331.99306.idtracker@ietfa.amsl.com> Date: Tue, 18 Jun 2013 14:59:38 -0700 X-Mailman-Approved-At: Wed, 19 Jun 2013 02:54:24 -0700 Cc: nsc@ietf.org, adrian@olddog.co.uk Subject: [nsc] New Non-WG Mailing List: nsc -- Network Service Chaining X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 18 Jun 2013 21:59:39 -0000 A new IETF non-working group email list has been created. List address: nsc@ietf.org Archive: http://www.ietf.org/mail-archive/web/nsc To subscribe: https://www.ietf.org/mailman/listinfo/nsc Purpose: Network services are widely deployed and essential in many network= s. The services provide a range of functions such as security, WAN acceleration, a= nd server load balancing. Service functions that form part of the overall serv= ice may be physically located at different points in the network infrastructure= such as the wide area network, data center, campus, and so forth. = New data center network and Internet cloud architectures require more flexi= ble network service deployment models. Additionally, the transition to virtual platforms requires an agile service insertion model that supports elastic service delivery, the movement of service functions and application workloa= ds in the network and the ability to easily bind service policy to granular information such as per-subscriber state. = Service chaining is a broad term used to describe a common model for delive= ring multiple services in a specific order. Service chaining de-couples service delivery from the underlying network topology and creates a dynamic services plane that addresses the requirements of cloud and virtual application deli= very. Traffic that requires service chaining is classified, and context is shared between the network and the services. = This list is for discussion of aspects of Network Service Chaining (NSC) th= at impact upon the IETF's work, the applicability of IETF protocols to NSC, as= well as new protocols and changes to IETF protocols that might be required. = A BoF has been proposed for NSC in Berlin. This list also serves to discuss= that BoF proposal. For additional information, please contact the list administrators. From adrian@olddog.co.uk Wed Jun 19 02:58:44 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id DF91B21F9F9E for ; Wed, 19 Jun 2013 02:58:44 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.587 X-Spam-Level: X-Spam-Status: No, score=-2.587 tagged_above=-999 required=5 tests=[AWL=0.012, BAYES_00=-2.599] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id yOrIl-JWE9cj for ; Wed, 19 Jun 2013 02:58:39 -0700 (PDT) Received: from asmtp3.iomartmail.com (asmtp3.iomartmail.com [62.128.201.159]) by ietfa.amsl.com (Postfix) with ESMTP id BCBED21F9F9C for ; Wed, 19 Jun 2013 02:58:38 -0700 (PDT) Received: from asmtp3.iomartmail.com (localhost.localdomain [127.0.0.1]) by asmtp3.iomartmail.com (8.13.8/8.13.8) with ESMTP id r5J9wbRW026429 for ; Wed, 19 Jun 2013 10:58:37 +0100 Received: from 950129200 (dsl-sp-81-140-15-32.in-addr.broadbandscope.com [81.140.15.32]) (authenticated bits=0) by asmtp3.iomartmail.com (8.13.8/8.13.8) with ESMTP id r5J9waZr026412 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO) for ; Wed, 19 Jun 2013 10:58:36 +0100 From: "Adrian Farrel" To: Date: Wed, 19 Jun 2013 10:58:35 +0100 Message-ID: <05cf01ce6cd3$90a7ac10$b1f70430$@olddog.co.uk> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Mailer: Microsoft Outlook 14.0 Thread-Index: Ac5s048h6TfwAJOOSDqZ41f+7z3ZbA== Content-Language: en-gb Subject: [nsc] NSC mailing list test message (please ignore) X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list Reply-To: adrian@olddog.co.uk List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 19 Jun 2013 09:58:45 -0000 From paulq@cisco.com Wed Jun 19 14:44:41 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 8C0B321E80B2 for ; Wed, 19 Jun 2013 14:44:41 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.449 X-Spam-Level: X-Spam-Status: No, score=-10.449 tagged_above=-999 required=5 tests=[AWL=0.150, BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id hz+poAlO9qAt for ; Wed, 19 Jun 2013 14:44:36 -0700 (PDT) Received: from mtv-iport-1.cisco.com (mtv-iport-1.cisco.com [173.36.130.12]) by ietfa.amsl.com (Postfix) with ESMTP id 7B09721E8097 for ; Wed, 19 Jun 2013 14:44:36 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=597; q=dns/txt; s=iport; t=1371678276; x=1372887876; h=from:content-transfer-encoding:subject:message-id:date: to:mime-version; bh=1gkgHAWOXv+IR12vDMryOrRgZrjWBW38xq+GGxATeKE=; b=UxCjxr0R7HwraRFEHokvdEXhEvsDJXsOuc3OyDs7e4cpejqyQNvoVv1a hy/FjZJUWbODUhLStX8SxWMYtkksQdGBbPsG2oqUPnmsTTyOxmxabsQ5q GR/oyGwlrYHWqb6uco1WJlnl/265g1XYCiDX0tlTwuDw1q5NhddXcs2cy c=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AlIHAPMlwlGrRDoJ/2dsb2JhbABagwkywGMWbQeCZIF9iCCbVaA5kklhA4kgjiGGIYsjgy8c X-IronPort-AV: E=Sophos;i="4.87,900,1363132800"; d="scan'208";a="80834552" Received: from mtv-core-4.cisco.com ([171.68.58.9]) by mtv-iport-1.cisco.com with ESMTP; 19 Jun 2013 21:44:33 +0000 Received: from [10.155.158.223] ([10.155.158.223]) by mtv-core-4.cisco.com (8.14.5/8.14.5) with ESMTP id r5JLiWIZ028432 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO) for ; Wed, 19 Jun 2013 21:44:32 GMT From: Paul Quinn Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: quoted-printable Message-Id: <18B55B0F-6D0E-4CE5-8C97-755BFE57951B@cisco.com> Date: Wed, 19 Jun 2013 14:44:32 -0700 To: "nsc@ietf.org" Mime-Version: 1.0 (Mac OS X Mail 6.5 \(1508\)) X-Mailer: Apple Mail (2.1508) Subject: [nsc] Prior to Berlin: some questions X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 19 Jun 2013 21:44:41 -0000 Fellow NSCers, In preparation for Berlin, I'd like to kick of some discussion on the = list: 1. How do you feel about the topic of service chaining in general? = There's a lot of discussion in the industry and amongst operators about = it, but on this list, what is the general thought about how service = chaining will impact networks? 2. I'm hoping folks have started to read the documents posted with the = BoF request. Please send you comments to the list. In addition to = those documents, what else do you expect to see addressed by this = effort? =20 Thanks Paul= From kgray@juniper.net Wed Jun 19 19:02:15 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 1CF9321F9F84 for ; Wed, 19 Jun 2013 19:02:15 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -1.467 X-Spam-Level: X-Spam-Status: No, score=-1.467 tagged_above=-999 required=5 tests=[AWL=2.000, BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4, UNRESOLVED_TEMPLATE=3.132] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Gf4MVa1pzjCP for ; Wed, 19 Jun 2013 19:01:50 -0700 (PDT) Received: from co9outboundpool.messaging.microsoft.com (co9ehsobe002.messaging.microsoft.com [207.46.163.25]) by ietfa.amsl.com (Postfix) with ESMTP id 2547B21F995C for ; Wed, 19 Jun 2013 19:01:46 -0700 (PDT) Received: from mail113-co9-R.bigfish.com (10.236.132.243) by CO9EHSOBE022.bigfish.com (10.236.130.85) with Microsoft SMTP Server id 14.1.225.23; Thu, 20 Jun 2013 02:01:38 +0000 Received: from mail113-co9 (localhost [127.0.0.1]) by mail113-co9-R.bigfish.com (Postfix) with ESMTP id 4B739A804AF for ; Thu, 20 Jun 2013 02:01:38 +0000 (UTC) X-Forefront-Antispam-Report: CIP:66.129.224.54; KIP:(null); UIP:(null); IPV:NLI; H:P-EMHUB02-HQ.jnpr.net; RD:none; EFVD:NLI X-SpamScore: -24 X-BigFish: PS-24(zzbb2dI98dI9371I1432Izz1f42h1ee6h1de0h1fdah1202h1e76h1d1ah1d2ah1fc6hzz1033IL8275bh8275dhz2fh2a8h683h839h944he5bhf0ah1220h1288h12a5h12a9h12bdh137ah13b6h1441h1504h1537h153bh162dh1631h1758h18e1h1946h19b5h1ad9h1b0ah1d0ch1d2eh1d3fh1dc1h1dfeh1dffh1e1dh1155h) Received-SPF: pass (mail113-co9: domain of juniper.net designates 66.129.224.54 as permitted sender) client-ip=66.129.224.54; envelope-from=kgray@juniper.net; helo=P-EMHUB02-HQ.jnpr.net ; -HQ.jnpr.net ; X-Forefront-Antispam-Report-Untrusted: CIP:157.56.237.149; KIP:(null); UIP:(null); (null); H:BY2PRD0511HT002.namprd05.prod.outlook.com; R:internal; EFV:INT Received: from mail113-co9 (localhost.localdomain [127.0.0.1]) by mail113-co9 (MessageSwitch) id 1371693693544367_7871; Thu, 20 Jun 2013 02:01:33 +0000 (UTC) Received: from CO9EHSMHS002.bigfish.com (unknown [10.236.132.242]) by mail113-co9.bigfish.com (Postfix) with ESMTP id 7DFC074005F for ; Thu, 20 Jun 2013 02:01:33 +0000 (UTC) Received: from P-EMHUB02-HQ.jnpr.net (66.129.224.54) by CO9EHSMHS002.bigfish.com (10.236.130.12) with Microsoft SMTP Server (TLS) id 14.1.225.23; Thu, 20 Jun 2013 02:01:33 +0000 Received: from P-CLDFE01-HQ.jnpr.net (172.24.192.59) by P-EMHUB02-HQ.jnpr.net (172.24.192.36) with Microsoft SMTP Server (TLS) id 8.3.213.0; Wed, 19 Jun 2013 19:01:32 -0700 Received: from o365mail.juniper.net (207.17.137.224) by o365mail.juniper.net (172.24.192.59) with Microsoft SMTP Server id 14.1.355.2; Wed, 19 Jun 2013 19:01:32 -0700 Received: from CO9EHSOBE037.bigfish.com (207.46.163.26) by o365mail.juniper.net (207.17.137.224) with Microsoft SMTP Server (TLS) id 14.1.355.2; Wed, 19 Jun 2013 19:13:30 -0700 Received: from mail40-co9-R.bigfish.com (10.236.132.234) by CO9EHSOBE037.bigfish.com (10.236.130.100) with Microsoft SMTP Server id 14.1.225.23; Thu, 20 Jun 2013 02:01:31 +0000 Received: from mail40-co9 (localhost [127.0.0.1]) by mail40-co9-R.bigfish.com (Postfix) with ESMTP id 62FCD100339 for ; Thu, 20 Jun 2013 02:01:31 +0000 (UTC) Received: from mail40-co9 (localhost.localdomain [127.0.0.1]) by mail40-co9 (MessageSwitch) id 1371693689750477_27039; Thu, 20 Jun 2013 02:01:29 +0000 (UTC) Received: from CO9EHSMHS021.bigfish.com (unknown [10.236.132.229]) by mail40-co9.bigfish.com (Postfix) with ESMTP id AAC407C0052; Thu, 20 Jun 2013 02:01:29 +0000 (UTC) Received: from BY2PRD0511HT002.namprd05.prod.outlook.com (157.56.237.149) by CO9EHSMHS021.bigfish.com (10.236.130.31) with Microsoft SMTP Server (TLS) id 14.1.225.23; Thu, 20 Jun 2013 02:01:29 +0000 Received: from BY2PRD0511MB403.namprd05.prod.outlook.com ([169.254.4.32]) by BY2PRD0511HT002.namprd05.prod.outlook.com ([10.255.129.37]) with mapi id 14.16.0324.000; Thu, 20 Jun 2013 02:01:28 +0000 From: Ken Gray To: Paul Quinn , "nsc@ietf.org" Thread-Topic: draft-quinn-nsh questions Thread-Index: AQHObVoSAzKwR2kPl0+DNR9Yu5OP1w== Date: Thu, 20 Jun 2013 02:01:27 +0000 Message-ID: In-Reply-To: <18B55B0F-6D0E-4CE5-8C97-755BFE57951B@cisco.com> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.14.0.111121 x-originating-ip: [10.255.129.4] Content-Type: text/plain; charset="us-ascii" Content-ID: <9D0E41CD2F69EA409BA5A2AFB6ADFA0F@namprd05.prod.outlook.com> Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-FOPE-CONNECTOR: Id%0$Dn%*$RO%0$TLS%0$FQDN%$TlsDn% X-FOPE-CONNECTOR: Id%12219$Dn%CISCO.COM$RO%2$TLS%5$FQDN%onpremiseedge-1018244.customer.frontbridge.com$TlsDn%o365mail.juniper.net X-FOPE-CONNECTOR: Id%12219$Dn%IETF.ORG$RO%2$TLS%5$FQDN%onpremiseedge-1018244.customer.frontbridge.com$TlsDn%o365mail.juniper.net X-OriginatorOrg: juniper.net Subject: [nsc] draft-quinn-nsh questions X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 02:02:16 -0000 Okay, out here in the open then ... For the header doc, I had two questions ... 1) The requirement for the context headers to be present if they're unused. Do you have a feel for what the reaction to the 128 bit overhead is going to be? If the context could be collapsed when unused, it could cushion the blow of the overhead, even though it may complicate header offset setting in machine microcode - I mean, it's a tradeoff. Interested in your thoughts there ... 2) I love the idea of the proxy. It applies to already existing virtual or non-virtual appliances so allows us to suck in vApps that were designed for other chaining schemes (and for a while this is going to be the wild-wild west), even proprietary ones. So, is the vision that this be a user space agent that is just "canned" as far as it's processing goes (like a utility function) - that is, it's a vendor-independent piece of software (potentially open-sourced) and/or a similar thing but with hooks to vendor specific overlay mapping software to help with the encap/decap? I mean, do I have it wrong or would the expectation be that this manipulates no metadata as there is no real context sharing with non-compliant apps - or am I missing something and it is more complex/customizable? On 6/19/13 5:44 PM, "Paul Quinn" wrote: >Fellow NSCers, > >In preparation for Berlin, I'd like to kick of some discussion on the >list: > >1. How do you feel about the topic of service chaining in general? >There's a lot of discussion in the industry and amongst operators about >it, but on this list, what is the general thought about how service >chaining will impact networks? > >2. I'm hoping folks have started to read the documents posted with the >BoF request. Please send you comments to the list. In addition to those >documents, what else do you expect to see addressed by this effort? > >Thanks >Paul >_______________________________________________ >nsc mailing list >nsc@ietf.org >https://www.ietf.org/mailman/listinfo/nsc > > From jmh@joelhalpern.com Wed Jun 19 19:12:46 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 098B621F9FBA for ; Wed, 19 Jun 2013 19:12:46 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -102.314 X-Spam-Level: X-Spam-Status: No, score=-102.314 tagged_above=-999 required=5 tests=[AWL=0.285, BAYES_00=-2.599, USER_IN_WHITELIST=-100] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id eKTBl1+sc57K for ; Wed, 19 Jun 2013 19:12:41 -0700 (PDT) Received: from mailc2.tigertech.net (mailc2.tigertech.net [208.80.4.156]) by ietfa.amsl.com (Postfix) with ESMTP id 6043D21F9FB7 for ; Wed, 19 Jun 2013 19:12:41 -0700 (PDT) Received: from localhost (localhost [127.0.0.1]) by mailc2.tigertech.net (Postfix) with ESMTP id 4F5F31A5908 for ; Wed, 19 Jun 2013 19:12:41 -0700 (PDT) X-Virus-Scanned: Debian amavisd-new at c2.tigertech.net Received: from [10.10.10.104] (pool-70-106-134-231.clppva.east.verizon.net [70.106.134.231]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by mailc2.tigertech.net (Postfix) with ESMTPSA id BF78B1A5902 for ; Wed, 19 Jun 2013 19:12:40 -0700 (PDT) Message-ID: <51C2650B.8040502@joelhalpern.com> Date: Wed, 19 Jun 2013 22:12:27 -0400 From: "Joel M. Halpern" User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:17.0) Gecko/20130509 Thunderbird/17.0.6 MIME-Version: 1.0 To: "nsc@ietf.org" References: In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: Re: [nsc] draft-quinn-nsh questions X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 02:12:46 -0000 Service Chaining is clearly a topic of interest to man of us individually and many of our companies either as vendors, service providers, or users of services. That's the easy part. I find myself concerned seeing a number of proposals for specific protocol mechanism when there is nothing resembling agreement on the needs, scales, or architecture for the problem. The number of sets of non-matching assumptions out there is large enough that I am having difficulty getting my head around what a working group would do. Defining protocol seems an odd place to start. Yours, Joel On 6/19/2013 10:01 PM, Ken Gray wrote: > Okay, out here in the open then ... > > For the header doc, I had two questions ... > > 1) The requirement for the context headers to be present if > they're unused. Do you have a feel for what the reaction to the 128 bit > overhead is going to be? If the context could be collapsed when > unused, it could cushion the blow of the overhead, even though it may > complicate header offset setting in machine microcode - I mean, it's a > tradeoff. Interested in your thoughts there ... > > > 2) I love the idea of the proxy. It applies to already > existing virtual or non-virtual appliances so allows us to suck in vApps > that were designed for other chaining schemes (and for a while this is > going to be the wild-wild west), even proprietary ones. > So, is the vision that this be a user space agent that is just "canned" > as far as it's processing goes (like a utility function) - that is, it's > a vendor-independent piece of software (potentially open-sourced) > and/or a similar thing but with hooks to vendor specific overlay mapping > software to help with the encap/decap? I mean, do I have it wrong or > would the expectation be that this manipulates no metadata as there is > no real context sharing with non-compliant apps - or am I missing > something and it is more complex/customizable? > > > On 6/19/13 5:44 PM, "Paul Quinn" wrote: > >> Fellow NSCers, >> >> In preparation for Berlin, I'd like to kick of some discussion on the >> list: >> >> 1. How do you feel about the topic of service chaining in general? >> There's a lot of discussion in the industry and amongst operators about >> it, but on this list, what is the general thought about how service >> chaining will impact networks? >> >> 2. I'm hoping folks have started to read the documents posted with the >> BoF request. Please send you comments to the list. In addition to those >> documents, what else do you expect to see addressed by this effort? >> >> Thanks >> Paul >> _______________________________________________ >> nsc mailing list >> nsc@ietf.org >> https://www.ietf.org/mailman/listinfo/nsc >> >> > > > > _______________________________________________ > nsc mailing list > nsc@ietf.org > https://www.ietf.org/mailman/listinfo/nsc > From adrian@olddog.co.uk Thu Jun 20 04:23:29 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id E752621F9ADC for ; Thu, 20 Jun 2013 04:23:29 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.452 X-Spam-Level: X-Spam-Status: No, score=-2.452 tagged_above=-999 required=5 tests=[AWL=0.147, BAYES_00=-2.599] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id FBKcOHgxz+oS for ; Thu, 20 Jun 2013 04:23:23 -0700 (PDT) Received: from asmtp3.iomartmail.com (asmtp3.iomartmail.com [62.128.201.159]) by ietfa.amsl.com (Postfix) with ESMTP id DA5A421F9ADA for ; Thu, 20 Jun 2013 04:23:21 -0700 (PDT) Received: from asmtp3.iomartmail.com (localhost.localdomain [127.0.0.1]) by asmtp3.iomartmail.com (8.13.8/8.13.8) with ESMTP id r5KBNKTt005442 for ; Thu, 20 Jun 2013 12:23:20 +0100 Received: from 950129200 (dsl-sp-81-140-15-32.in-addr.broadbandscope.com [81.140.15.32]) (authenticated bits=0) by asmtp3.iomartmail.com (8.13.8/8.13.8) with ESMTP id r5KBNJwP005434 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO) for ; Thu, 20 Jun 2013 12:23:20 +0100 From: "Adrian Farrel" To: Date: Thu, 20 Jun 2013 12:23:17 +0100 Message-ID: <07c001ce6da8$904f9130$b0eeb390$@olddog.co.uk> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Mailer: Microsoft Outlook 14.0 Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywew== Content-Language: en-gb Subject: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list Reply-To: adrian@olddog.co.uk List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 11:23:30 -0000 Hi, I know this cuts into use cases, but I think it will turn out to be important for clarity and scope. There is a risk that "we all know what we mean", but that no-one clearly articulates what "network service chaining" is. (It is a little like NFV where the definition also seems to be hard to pin down.) I don't want to rat-hole too much on definitions, but it would help if we can draw some lines. When is "service chaining" not "network service chaining"? Thanks, Adrian From mohamed.boucadair@orange.com Thu Jun 20 04:28:16 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 8ABED21F9B65 for ; Thu, 20 Jun 2013 04:28:16 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.195 X-Spam-Level: X-Spam-Status: No, score=-2.195 tagged_above=-999 required=5 tests=[AWL=0.052, BAYES_00=-2.599, HELO_EQ_FR=0.35, HTML_MESSAGE=0.001, UNPARSEABLE_RELAY=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id i2zIrpkOnuYx for ; Thu, 20 Jun 2013 04:28:12 -0700 (PDT) Received: from relais-inet.francetelecom.com (relais-ias91.francetelecom.com [193.251.215.91]) by ietfa.amsl.com (Postfix) with ESMTP id 00B0C21F9B5E for ; Thu, 20 Jun 2013 04:28:11 -0700 (PDT) Received: from omfedm05.si.francetelecom.fr (unknown [xx.xx.xx.1]) by omfedm13.si.francetelecom.fr (ESMTP service) with ESMTP id B2296325BED for ; Thu, 20 Jun 2013 13:28:10 +0200 (CEST) Received: from PUEXCH81.nanterre.francetelecom.fr (unknown [10.101.44.34]) by omfedm05.si.francetelecom.fr (ESMTP service) with ESMTP id 97E3C35C045 for ; Thu, 20 Jun 2013 13:28:10 +0200 (CEST) Received: from PUEXCB1B.nanterre.francetelecom.fr ([10.101.44.13]) by PUEXCH81.nanterre.francetelecom.fr ([10.101.44.34]) with mapi; Thu, 20 Jun 2013 13:28:10 +0200 From: To: "nsc@ietf.org" Date: Thu, 20 Jun 2013 13:28:09 +0200 Thread-Topic: NLFC: Use Cases Thread-Index: Ac5tqTz2aym2EOlASR6C0afVpw7eRQ== Message-ID: <94C682931C08B048B7A8645303FDC9F36EDB2591F9@PUEXCB1B.nanterre.francetelecom.fr> Accept-Language: fr-FR Content-Language: fr-FR X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: fr-FR Content-Type: multipart/alternative; boundary="_000_94C682931C08B048B7A8645303FDC9F36EDB2591F9PUEXCB1Bnante_" MIME-Version: 1.0 X-PMX-Version: 5.6.1.2065439, Antispam-Engine: 2.7.2.376379, Antispam-Data: 2013.6.20.62421 Subject: [nsc] NLFC: Use Cases X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 11:28:16 -0000 --_000_94C682931C08B048B7A8645303FDC9F36EDB2591F9PUEXCB1Bnante_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Dear all, It would be useful if we all share the use cases that motivate our interest= on this work. The effort documented in http://tools.ietf.org/html/draft-boucadair-network= -function-chaining-00 is mainly motivated by the current work we are invest= igating to rationalize the Gi interface (mobile context) in the hope to sim= ply the complexity encountered in current deployments. Indeed, * Plenty of functions are enabled in that segment * Lack of visibility on dependency between these functions * Complexity to introduce new function because of the constraint on underly= ing topology * Lack of automated means to assess the impact of withdrawing a function fr= om the path * The connectivity service logic may be frozen because of the dependency on= the physical topology Because we are aware these issues are not specific to our mobile use case, = we tried to generalize the motivations space as follows: o Efficiently master the chained activation of network-located functions, regardless of the underlying network topology and routing policies. o Allow for differentiated packet forwarding by selecting the set of network-located functions to be invoked. o Allow to easily change the chronology of the activation of network-located functions to be invoked. o Allow to easily change the set of network-located functions to be invoked. o Ease withdrawal of network-located functions and minimize any subsequent topology upgrade. o Automate the overall process of generating and enforcing policies to accommodate a set of network connectivity objectives. Cheers, Med --_000_94C682931C08B048B7A8645303FDC9F36EDB2591F9PUEXCB1Bnante_ Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable

Dear all,<= /span>

 

I= t would be useful if we all share the use cases that motivate our interest = on this work.

 

The effort documented in http://tools.iet= f.org/html/draft-boucadair-network-function-chaining-00 is mainly motiv= ated by the current work we are investigating to rationalize the Gi interfa= ce (mobile context) in the hope to simply the complexity encountered in cur= rent deployments. Indeed,

* Plenty = of functions are enabled in that segment

* Lack of visibility on dependency between these functions =

* Complexity to introduce new function because= of the constraint on underlying topology

* Lack of automated means to assess the impact of withdrawing a funct= ion from the path

* The connectivity= service logic may be frozen because of the dependency on the physical topo= logy

 

Because we are aware these issues are not specific to our mo= bile use case, we tried to generalize the motivations space as follows:

 

   o  Efficiently master the = chained activation of network-located

      functions, regardl= ess of the underlying network topology and

      routing polic= ies.

 &= nbsp; o  Allow for differentiated packet forwarding by selecting the s= et of

 =      network-located functions to be invoked.

   o  = Allow to easily change the chronology of the activation of

    &nbs= p; network-located functions to be invoked.

   o  Allow to easily chan= ge the set of network-located functions to be

      invoked.

   = o  Ease withdrawal of network-located functions and minimize any<= /o:p>

   = ;   subsequent topology upgrade.

   o  Automate the overall pr= ocess of generating and enforcing policies

      to accommodat= e a set of network connectivity objectives.

 

Cheers,<= /span>

Med

<= span lang=3DEN-US style=3D'font-size:10.0pt;font-family:"Courier New"'> 

= --_000_94C682931C08B048B7A8645303FDC9F36EDB2591F9PUEXCB1Bnante_-- From mohamed.boucadair@orange.com Thu Jun 20 05:00:21 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 1C2B521F9C0D for ; Thu, 20 Jun 2013 05:00:21 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.198 X-Spam-Level: X-Spam-Status: No, score=-2.198 tagged_above=-999 required=5 tests=[AWL=0.050, BAYES_00=-2.599, HELO_EQ_FR=0.35, UNPARSEABLE_RELAY=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 8-KEK43ZUjiv for ; Thu, 20 Jun 2013 05:00:17 -0700 (PDT) Received: from relais-inet.francetelecom.com (relais-ias91.francetelecom.com [193.251.215.91]) by ietfa.amsl.com (Postfix) with ESMTP id AA05021F9C10 for ; Thu, 20 Jun 2013 05:00:16 -0700 (PDT) Received: from omfedm05.si.francetelecom.fr (unknown [xx.xx.xx.1]) by omfedm12.si.francetelecom.fr (ESMTP service) with ESMTP id D66FA18D7B5; Thu, 20 Jun 2013 14:00:15 +0200 (CEST) Received: from PUEXCH21.nanterre.francetelecom.fr (unknown [10.101.44.28]) by omfedm05.si.francetelecom.fr (ESMTP service) with ESMTP id A28F135C097; Thu, 20 Jun 2013 14:00:15 +0200 (CEST) Received: from PUEXCB1B.nanterre.francetelecom.fr ([10.101.44.13]) by PUEXCH21.nanterre.francetelecom.fr ([10.101.44.28]) with mapi; Thu, 20 Jun 2013 14:00:15 +0200 From: To: "adrian@olddog.co.uk" , "nsc@ietf.org" Date: Thu, 20 Jun 2013 14:00:14 +0200 Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrew Message-ID: <94C682931C08B048B7A8645303FDC9F36EDB259223@PUEXCB1B.nanterre.francetelecom.fr> References: <07c001ce6da8$904f9130$b0eeb390$@olddog.co.uk> In-Reply-To: <07c001ce6da8$904f9130$b0eeb390$@olddog.co.uk> Accept-Language: fr-FR Content-Language: fr-FR X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: fr-FR Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-PMX-Version: 5.6.1.2065439, Antispam-Engine: 2.7.2.376379, Antispam-Data: 2013.5.21.113319 Cc: JACQUENET Christian OLNC/OLN Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 12:00:21 -0000 Hi Adrian,=20 You raised a good point.=20 I do think the use of "service" and even "network service" will induce conf= usion. "service" is an overloaded term which does not have the same signifi= cation for network engineers, application designs, marketing people, etc.=20 In addition, "service" is already a packaging of several functions. It may = implicitly require the invocation of various functions and the combinations= of multiple actions involving several (network) nodes. This is for instanc= e the notion of service as defined in RFC 2474. For those reasons, I would suggest the use of: * a non-overloaded term * a more neutral term which can suit various use cases of interest of this = group As an alternative, I'm proposing the use of NLF (Network-Located Function).= A proposed definition for NLF is provided below: o NLF (Network-Located Function): refers to a function which is enabled in the network operated by an administrative entity. NLF can be for example: firewall (e.g., [RFC6092]), DPI (Deep Packet Inspection), LI (Lawful Intercept) module, NAT44 [RFC3022], NAT64 [RFC6146], DS-Lite AFTR [RFC6333], NPTv6 [RFC6296], HOST_ID injection, HTTP Header Enrichment function, TCP optimizer, etc.=20 Cheers, Med >-----Message d'origine----- >De=A0: nsc-bounces@ietf.org [mailto:nsc-bounces@ietf.org] De la part de >Adrian Farrel >Envoy=E9=A0: jeudi 20 juin 2013 13:23 >=C0=A0: nsc@ietf.org >Objet=A0: [nsc] What is a service? Why is NSC not SC? > >Hi, > >I know this cuts into use cases, but I think it will turn out to be >important >for clarity and scope. > >There is a risk that "we all know what we mean", but that no-one clearly >articulates what "network service chaining" is. (It is a little like NFV >where >the definition also seems to be hard to pin down.) > >I don't want to rat-hole too much on definitions, but it would help if we >can >draw some lines. When is "service chaining" not "network service chaining"= ? > >Thanks, >Adrian > >_______________________________________________ >nsc mailing list >nsc@ietf.org >https://www.ietf.org/mailman/listinfo/nsc From kgray@juniper.net Thu Jun 20 05:55:14 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D9C1521F9C8F for ; Thu, 20 Jun 2013 05:55:13 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -0.467 X-Spam-Level: X-Spam-Status: No, score=-0.467 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_LOW=-1, UNRESOLVED_TEMPLATE=3.132] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id wV9LgWg4aw18 for ; Thu, 20 Jun 2013 05:55:08 -0700 (PDT) Received: from co1outboundpool.messaging.microsoft.com (co1ehsobe004.messaging.microsoft.com [216.32.180.187]) by ietfa.amsl.com (Postfix) with ESMTP id 0149521F9C71 for ; Thu, 20 Jun 2013 05:55:07 -0700 (PDT) Received: from mail35-co1-R.bigfish.com (10.243.78.239) by CO1EHSOBE025.bigfish.com (10.243.66.88) with Microsoft SMTP Server id 14.1.225.23; Thu, 20 Jun 2013 12:55:07 +0000 Received: from mail35-co1 (localhost [127.0.0.1]) by mail35-co1-R.bigfish.com (Postfix) with ESMTP id 2835A5C0423 for ; Thu, 20 Jun 2013 12:55:07 +0000 (UTC) X-Forefront-Antispam-Report: CIP:66.129.224.51; KIP:(null); UIP:(null); IPV:NLI; H:P-EMHUB01-HQ.jnpr.net; RD:none; EFVD:NLI X-SpamScore: -23 X-BigFish: PS-23(zz98dI9371I1432I4015Izz1f42h1ee6h1de0h1fdah1202h1e76h1d1ah1d2ah1fc6hz31iz1033IL8275dhz2fh2a8h683h839h944hd25he5bhf0ah1220h1288h12a5h12a9h12bdh137ah13b6h1441h1504h1537h153bh162dh1631h1758h18e1h1946h19b5h19ceh1ad9h1b0ah1d0ch1d2eh1d3fh1dfeh1dffh1e1dh1155h) Received-SPF: pass (mail35-co1: domain of juniper.net designates 66.129.224.51 as permitted sender) client-ip=66.129.224.51; envelope-from=kgray@juniper.net; helo=P-EMHUB01-HQ.jnpr.net ; -HQ.jnpr.net ; X-Forefront-Antispam-Report-Untrusted: CIP:157.56.237.149; KIP:(null); UIP:(null); (null); H:BY2PRD0511HT004.namprd05.prod.outlook.com; R:internal; EFV:INT Received: from mail35-co1 (localhost.localdomain [127.0.0.1]) by mail35-co1 (MessageSwitch) id 137173290519806_18762; Thu, 20 Jun 2013 12:55:05 +0000 (UTC) Received: from CO1EHSMHS017.bigfish.com (unknown [10.243.78.239]) by mail35-co1.bigfish.com (Postfix) with ESMTP id 027484004A for ; Thu, 20 Jun 2013 12:55:05 +0000 (UTC) Received: from P-EMHUB01-HQ.jnpr.net (66.129.224.51) by CO1EHSMHS017.bigfish.com (10.243.66.27) with Microsoft SMTP Server (TLS) id 14.1.225.23; Thu, 20 Jun 2013 12:55:04 +0000 Received: from P-CLDFE02-HQ.jnpr.net (172.24.192.60) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server (TLS) id 8.3.213.0; Thu, 20 Jun 2013 05:55:04 -0700 Received: from o365mail.juniper.net (207.17.137.224) by o365mail.juniper.net (172.24.192.60) with Microsoft SMTP Server id 14.1.355.2; Thu, 20 Jun 2013 05:55:03 -0700 Received: from tx2outboundpool.messaging.microsoft.com (65.55.88.13) by o365mail.juniper.net (207.17.137.224) with Microsoft SMTP Server (TLS) id 14.1.355.2; Thu, 20 Jun 2013 06:07:01 -0700 Received: from mail81-tx2-R.bigfish.com (10.9.14.253) by TX2EHSOBE009.bigfish.com (10.9.40.29) with Microsoft SMTP Server id 14.1.225.23; Thu, 20 Jun 2013 12:55:03 +0000 Received: from mail81-tx2 (localhost [127.0.0.1]) by mail81-tx2-R.bigfish.com (Postfix) with ESMTP id F056E60194 for ; Thu, 20 Jun 2013 12:55:02 +0000 (UTC) Received: from mail81-tx2 (localhost.localdomain [127.0.0.1]) by mail81-tx2 (MessageSwitch) id 137173290188166_5534; Thu, 20 Jun 2013 12:55:01 +0000 (UTC) Received: from TX2EHSMHS009.bigfish.com (unknown [10.9.14.244]) by mail81-tx2.bigfish.com (Postfix) with ESMTP id 0749732004C; Thu, 20 Jun 2013 12:55:01 +0000 (UTC) Received: from BY2PRD0511HT004.namprd05.prod.outlook.com (157.56.237.149) by TX2EHSMHS009.bigfish.com (10.9.99.109) with Microsoft SMTP Server (TLS) id 14.1.225.23; Thu, 20 Jun 2013 12:54:57 +0000 Received: from BY2PRD0511MB403.namprd05.prod.outlook.com ([169.254.4.32]) by BY2PRD0511HT004.namprd05.prod.outlook.com ([10.255.129.39]) with mapi id 14.16.0324.000; Thu, 20 Jun 2013 12:54:56 +0000 From: Ken Gray To: "" Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewADM8Ao Date: Thu, 20 Jun 2013 12:54:56 +0000 Message-ID: <9AB16734-A91E-4542-B073-7C83891EB65E@juniper.net> References: <07c001ce6da8$904f9130$b0eeb390$@olddog.co.uk> In-Reply-To: <07c001ce6da8$904f9130$b0eeb390$@olddog.co.uk> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [76.100.123.238] Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-FOPE-CONNECTOR: Id%0$Dn%*$RO%0$TLS%0$FQDN%$TlsDn% X-FOPE-CONNECTOR: Id%12219$Dn%OLDDOG.CO.UK$RO%2$TLS%5$FQDN%onpremiseedge-1018244.customer.frontbridge.com$TlsDn%o365mail.juniper.net X-FOPE-CONNECTOR: Id%12219$Dn%IETF.ORG$RO%2$TLS%5$FQDN%onpremiseedge-1018244.customer.frontbridge.com$TlsDn%o365mail.juniper.net X-OriginatorOrg: juniper.net Cc: "nsc@ietf.org" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 12:55:14 -0000 To me, if you are trying to make distinctions, I think its all in focus. W= e're concerned with how the datagram gets from point A to point B, where A = and B may be described as a service element (i think Jim's SEP terminology = is apt here) and that element may require either an imbedded or orchestrate= d clue/context (as required by a more global concept of service that may en= compass many elements. That delivery may be constrained by a number of network-related factors tha= t require the application of policy, knowledge of aspects of topology, and = continual monitoring/maintenance of those factors. So, the "network" focus is that delivery. Or to define in negatives, its not concerned with the decomposition of the = service or element, the optimization of the service platform or best practi= ces in creating, maintaining or administering a service (except for the are= as that ultimately end up manipulating the datagram itself or the forwardin= g logic of the element (eg the interface with orchestration). I know that spitball is incomplete. I don't think we need another way to s= ay "service" but should focus on the "network" component. Sent from my iPhone On Jun 20, 2013, at 7:23 AM, "Adrian Farrel" wrote: > Hi, >=20 > I know this cuts into use cases, but I think it will turn out to be impor= tant > for clarity and scope. >=20 > There is a risk that "we all know what we mean", but that no-one clearly > articulates what "network service chaining" is. (It is a little like NFV = where > the definition also seems to be hard to pin down.) >=20 > I don't want to rat-hole too much on definitions, but it would help if we= can > draw some lines. When is "service chaining" not "network service chaining= "? >=20 > Thanks, > Adrian >=20 > _______________________________________________ > nsc mailing list > nsc@ietf.org > https://www.ietf.org/mailman/listinfo/nsc >=20 >=20 From narten@us.ibm.com Thu Jun 20 06:52:10 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id DE6DA21F9BA7 for ; Thu, 20 Jun 2013 06:52:10 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -110.599 X-Spam-Level: X-Spam-Status: No, score=-110.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8, USER_IN_WHITELIST=-100] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id wvKqm7iJE2zU for ; Thu, 20 Jun 2013 06:52:03 -0700 (PDT) Received: from e7.ny.us.ibm.com (e7.ny.us.ibm.com [32.97.182.137]) by ietfa.amsl.com (Postfix) with ESMTP id E4B4B21F9BA4 for ; Thu, 20 Jun 2013 06:51:50 -0700 (PDT) Received: from /spool/local by e7.ny.us.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Thu, 20 Jun 2013 09:51:49 -0400 Received: from d01dlp01.pok.ibm.com (9.56.250.166) by e7.ny.us.ibm.com (192.168.1.107) with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted; Thu, 20 Jun 2013 09:51:46 -0400 Received: from d01relay05.pok.ibm.com (d01relay05.pok.ibm.com [9.56.227.237]) by d01dlp01.pok.ibm.com (Postfix) with ESMTP id 7AF6138C8042 for ; Thu, 20 Jun 2013 09:51:45 -0400 (EDT) Received: from d03av05.boulder.ibm.com (d03av05.boulder.ibm.com [9.17.195.85]) by d01relay05.pok.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id r5KDpjlc296778 for ; Thu, 20 Jun 2013 09:51:45 -0400 Received: from d03av05.boulder.ibm.com (loopback [127.0.0.1]) by d03av05.boulder.ibm.com (8.14.4/8.13.1/NCO v10.0 AVout) with ESMTP id r5KDphAf027826 for ; Thu, 20 Jun 2013 07:51:43 -0600 Received: from cichlid.raleigh.ibm.com (sig-9-76-159-14.mts.ibm.com [9.76.159.14]) by d03av05.boulder.ibm.com (8.14.4/8.13.1/NCO v10.0 AVin) with ESMTP id r5KDpeoH027658 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 20 Jun 2013 07:51:41 -0600 Received: from cichlid.raleigh.ibm.com (localhost [127.0.0.1]) by cichlid.raleigh.ibm.com (8.14.4/8.12.5) with ESMTP id r5KDoRET023472; Thu, 20 Jun 2013 09:50:28 -0400 Message-Id: <201306201350.r5KDoRET023472@cichlid.raleigh.ibm.com> To: "Joel M. Halpern" In-reply-to: <51C2650B.8040502@joelhalpern.com> References: <51C2650B.8040502@joelhalpern.com> Comments: In-reply-to "Joel M. Halpern" message dated "Wed, 19 Jun 2013 22:12:27 -0400." Date: Thu, 20 Jun 2013 09:50:27 -0400 From: Thomas Narten X-TM-AS-MML: No X-Content-Scanned: Fidelis XPS MAILER x-cbid: 13062013-5806-0000-0000-000021CC704B Cc: "nsc@ietf.org" Subject: Re: [nsc] draft-quinn-nsh questions X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 13:52:11 -0000 > Service Chaining is clearly a topic of interest to man of us > individually and many of our companies either as vendors, service > providers, or users of services. That's the easy part. Totally agree! > I find myself concerned seeing a number of proposals for specific > protocol mechanism when there is nothing resembling agreement on the > needs, scales, or architecture for the problem. The number of sets of > non-matching assumptions out there is large enough that I am having > difficulty getting my head around what a working group would do. > Defining protocol seems an odd place to start. Agree here as well. I suspect getting on the same page in terms of the problem statement will not be too hard. I agree with much of what is in the draft-quinn-nsc-problem-statement-00.txt. Saying where the pain is is relatively easy. But in terms of how best to address those issues, as in architectural choices, and what the tradeoffs are with different approaches, that will be harder. draft-quinn-nsh has a lot of fields that are more or less opaque or are just identifiers for some sort of context information that is not defined. That makes it challenging to understand what sort of general *architecture* is being thought of here. Thomas From dmm@1-4-5.net Thu Jun 20 07:07:54 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7F3F721F9A00 for ; Thu, 20 Jun 2013 07:07:54 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -0.814 X-Spam-Level: X-Spam-Status: No, score=-0.814 tagged_above=-999 required=5 tests=[AWL=-0.388, BAYES_00=-2.599, FH_RELAY_NODNS=1.451, FM_FORGED_GMAIL=0.622, RDNS_NONE=0.1] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id AqOddK4epsFG for ; Thu, 20 Jun 2013 07:07:50 -0700 (PDT) Received: from mail-oa0-x232.google.com (mail-oa0-x232.google.com [IPv6:2607:f8b0:4003:c02::232]) by ietfa.amsl.com (Postfix) with ESMTP id 5F1DB21F9A16 for ; Thu, 20 Jun 2013 07:07:50 -0700 (PDT) Received: by mail-oa0-f50.google.com with SMTP id k7so7859243oag.23 for ; Thu, 20 Jun 2013 07:07:49 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:x-originating-ip:in-reply-to:references:date :message-id:subject:from:to:cc:content-type:x-gm-message-state; bh=TdIgRQinyEo64Em6WNEnq0QNcsJL3NIvba8s/WPgpvg=; b=f4gAHXFqDlcp8XIaTCRo+tXclCparVi9XjDk9iqQ3iv7zXVMUAZVSrd7+wVMp/VUk2 GwCpQHnbXVCjByE00MHGV9T+rAkFrpiW41H7Ntj9XctDHVdvGeIgykTZhVu4vDMaO5oK +cucfAaqGj2caM6o0cPiSKd7mn44KoxFVXwyXJFwFCA3pqzx3Q5uf6o8M575EBigO7HM 7QUGeyuKGShdGcjEQarw5LZby05Qr64uVstmcnMWn2hj9R6HXFb7dM5IPttRfHRoWT4e lzVcDx5mg8w0CL2teKzu5A+asnuusry1gM0XuW4muWceuFcclzCLSSRT0el1zHJ4t3IS zKZg== MIME-Version: 1.0 X-Received: by 10.182.220.199 with SMTP id py7mr1523828obc.92.1371737269771; Thu, 20 Jun 2013 07:07:49 -0700 (PDT) Received: by 10.182.31.68 with HTTP; Thu, 20 Jun 2013 07:07:49 -0700 (PDT) X-Originating-IP: [128.223.156.253] In-Reply-To: <51C2650B.8040502@joelhalpern.com> References: <51C2650B.8040502@joelhalpern.com> Date: Thu, 20 Jun 2013 07:07:49 -0700 Message-ID: From: David Meyer To: "Joel M. Halpern" Content-Type: text/plain; charset=ISO-8859-1 X-Gm-Message-State: ALoCoQlWtKCYqW3myr1CZ5lrWZ4mllDY5GdyyyMYPfmON1roI7ggHxTAIc5Hl15z6NGmMxJShlmy Cc: "nsc@ietf.org" Subject: Re: [nsc] draft-quinn-nsh questions X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 14:07:54 -0000 On Wed, Jun 19, 2013 at 7:12 PM, Joel M. Halpern wrote: > Service Chaining is clearly a topic of interest to man of us individually > and many of our companies either as vendors, service providers, or users of > services. That's the easy part. Agree Joel. > > I find myself concerned seeing a number of proposals for specific protocol > mechanism when there is nothing resembling agreement on the needs, scales, > or architecture for the problem. The number of sets of non-matching > assumptions out there is large enough that I am having difficulty getting my > head around what a working group would do. Defining protocol seems an odd > place to start. Again, agreed. --dmm > > Yours, > Joel > > > On 6/19/2013 10:01 PM, Ken Gray wrote: >> >> Okay, out here in the open then ... >> >> For the header doc, I had two questions ... >> >> 1) The requirement for the context headers to be present if >> they're unused. Do you have a feel for what the reaction to the 128 bit >> overhead is going to be? If the context could be collapsed when >> unused, it could cushion the blow of the overhead, even though it may >> complicate header offset setting in machine microcode - I mean, it's a >> tradeoff. Interested in your thoughts there ... >> >> >> 2) I love the idea of the proxy. It applies to already >> existing virtual or non-virtual appliances so allows us to suck in vApps >> that were designed for other chaining schemes (and for a while this is >> going to be the wild-wild west), even proprietary ones. >> So, is the vision that this be a user space agent that is just "canned" >> as far as it's processing goes (like a utility function) - that is, it's >> a vendor-independent piece of software (potentially open-sourced) >> and/or a similar thing but with hooks to vendor specific overlay mapping >> software to help with the encap/decap? I mean, do I have it wrong or >> would the expectation be that this manipulates no metadata as there is >> no real context sharing with non-compliant apps - or am I missing >> something and it is more complex/customizable? >> >> >> On 6/19/13 5:44 PM, "Paul Quinn" wrote: >> >>> Fellow NSCers, >>> >>> In preparation for Berlin, I'd like to kick of some discussion on the >>> list: >>> >>> 1. How do you feel about the topic of service chaining in general? >>> There's a lot of discussion in the industry and amongst operators about >>> it, but on this list, what is the general thought about how service >>> chaining will impact networks? >>> >>> 2. I'm hoping folks have started to read the documents posted with the >>> BoF request. Please send you comments to the list. In addition to those >>> documents, what else do you expect to see addressed by this effort? >>> >>> Thanks >>> Paul >>> _______________________________________________ >>> nsc mailing list >>> nsc@ietf.org >>> https://www.ietf.org/mailman/listinfo/nsc >>> >>> >> >> >> >> _______________________________________________ >> nsc mailing list >> nsc@ietf.org >> https://www.ietf.org/mailman/listinfo/nsc >> > _______________________________________________ > nsc mailing list > nsc@ietf.org > https://www.ietf.org/mailman/listinfo/nsc From adrian@olddog.co.uk Thu Jun 20 07:27:21 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6539B21F9BA6 for ; Thu, 20 Jun 2013 07:27:21 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.417 X-Spam-Level: X-Spam-Status: No, score=-2.417 tagged_above=-999 required=5 tests=[AWL=0.182, BAYES_00=-2.599] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id fzYDM-nXEBK4 for ; Thu, 20 Jun 2013 07:27:15 -0700 (PDT) Received: from asmtp2.iomartmail.com (asmtp2.iomartmail.com [62.128.201.249]) by ietfa.amsl.com (Postfix) with ESMTP id 434C021F9C58 for ; Thu, 20 Jun 2013 07:27:02 -0700 (PDT) Received: from asmtp2.iomartmail.com (localhost.localdomain [127.0.0.1]) by asmtp2.iomartmail.com (8.13.8/8.13.8) with ESMTP id r5KEQxl5029269 for ; Thu, 20 Jun 2013 15:27:00 +0100 Received: from 950129200 (dsl-sp-81-140-15-32.in-addr.broadbandscope.com [81.140.15.32]) (authenticated bits=0) by asmtp2.iomartmail.com (8.13.8/8.13.8) with ESMTP id r5KEQw9K029226 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO) for ; Thu, 20 Jun 2013 15:26:59 +0100 From: "Adrian Farrel" To: References: <07c001ce6da8$904f9130$b0eeb390$@olddog.co.uk> <94C682931C08B048B7A8645303FDC9F36EDB259223@PUEXCB1B.nanterre.francetelecom.fr> In-Reply-To: <94C682931C08B048B7A8645303FDC9F36EDB259223@PUEXCB1B.nanterre.francetelecom.fr> Date: Thu, 20 Jun 2013 15:26:57 +0100 Message-ID: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Mailer: Microsoft Outlook 14.0 Thread-Index: AQHK3KVnq8c9UrscD5I8tUEuvfdA4QOO3MLWmSlTBWA= Content-Language: en-gb Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list Reply-To: adrian@olddog.co.uk List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 14:27:21 -0000 Thanks to Ken and Med for replies. Choosing to respond on Med's thread because he gave... > o NLF (Network-Located Function): refers to a function which is > enabled in the network operated by an administrative entity. NLF > can be for example: firewall (e.g., [RFC6092]), DPI (Deep Packet > Inspection), LI (Lawful Intercept) module, NAT44 [RFC3022], NAT64 > [RFC6146], DS-Lite AFTR [RFC6333], NPTv6 [RFC6296], HOST_ID > injection, HTTP Header Enrichment function, TCP optimizer, etc. Something I pick out of this is that all the example functions are packet-oriented functions (or possibly flow-oriented) and not application layer services. I think this is significant and not accidental :-) Can we capture that idea in a definition? I note that it is hinted at in Paul and Jim's BoF proposal text and it is there because I asked them why this proposal was aimed at the Routing Area not the Applications Area. To which the answer was "because we want to steer packets or flows to the place where the next service will be applied." Adrian From jguichar@cisco.com Thu Jun 20 07:59:50 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 8C00221F9BD4 for ; Thu, 20 Jun 2013 07:59:50 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.599 X-Spam-Level: X-Spam-Status: No, score=-10.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id PkJfU3keAzrA for ; Thu, 20 Jun 2013 07:59:45 -0700 (PDT) Received: from rcdn-iport-8.cisco.com (rcdn-iport-8.cisco.com [173.37.86.79]) by ietfa.amsl.com (Postfix) with ESMTP id A772D21F9BD5 for ; Thu, 20 Jun 2013 07:59:35 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=6802; q=dns/txt; s=iport; t=1371740375; x=1372949975; h=from:to:subject:date:message-id:in-reply-to:mime-version; bh=svwCxjOL3F2ZcjTeYKgH8sg6NhbkluxjIW0wjiJZFRo=; b=czX9cbcWEWSaJ66c+bYXwXH+n2/NJKzx1OPsO1Eet2PiCf37oJedo8NK TRxG9/vgjR1/DIF7CTT58DK6mu917fgIaDA2pRs8p3LZoYroepbjK1yFC dWOmP9yI/SBciez7veSpFZi+OUFrusLpgp4hSP4thRC/94mVO1uSiBLJ5 w=; X-Files: default.xml : 3222 X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AggFAA4Yw1GtJV2d/2dsb2JhbABbgwkxSb4sgQp8FnSCJQEEAQEBax0BCCImJQslAQEEARIIBogADLtiBI4GF3s4gwBhA5ADgSyQQ4cVgw+BaT8 X-IronPort-AV: E=Sophos;i="4.87,905,1363132800"; d="rels'?xml'?scan'72,48,208";a="225351850" Received: from rcdn-core-6.cisco.com ([173.37.93.157]) by rcdn-iport-8.cisco.com with ESMTP; 20 Jun 2013 14:59:35 +0000 Received: from xhc-aln-x02.cisco.com (xhc-aln-x02.cisco.com [173.36.12.76]) by rcdn-core-6.cisco.com (8.14.5/8.14.5) with ESMTP id r5KExZM9022108 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Thu, 20 Jun 2013 14:59:35 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.79]) by xhc-aln-x02.cisco.com ([173.36.12.76]) with mapi id 14.02.0318.004; Thu, 20 Jun 2013 09:59:34 -0500 From: "Jim Guichard (jguichar)" To: "adrian@olddog.co.uk" , "nsc@ietf.org" Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAJpiAA Date: Thu, 20 Jun 2013 14:59:34 +0000 Message-ID: <68B171751455884590F8E38E96416F363F53F9F7@xmb-rcd-x01.cisco.com> In-Reply-To: <07c001ce6da8$904f9130$b0eeb390$@olddog.co.uk> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: yes X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.2.3.120616 x-originating-ip: [10.98.130.212] Content-Type: multipart/mixed; boundary="_002_68B171751455884590F8E38E96416F363F53F9F7xmbrcdx01ciscoc_" MIME-Version: 1.0 Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 14:59:50 -0000 --_002_68B171751455884590F8E38E96416F363F53F9F7xmbrcdx01ciscoc_ Content-Type: text/plain; charset="us-ascii" Content-ID: Content-Transfer-Encoding: quoted-printable Hi Adrian,=20 >From my perspective "service" is an overloaded term, so we attempted to bound the concept of service. Depending on perspective a service could be an individual vertical service such as a financial or sales application, a transit service such as network security or IP connectivity, a service providing compute or storage resources, a network-based service such as a firewall, a particular piece of content such as a video stream, or some combination of all of these. By adding the term network, we imply that we are focused on functions that are linked via the network, and have network-relevant attributes (e.g. network location). In addition to this some services will require metadata and some will not; for those that do we need a way to pass said metadata between service instances and this is where additional mechanisms above and beyond "service routing" come into play such as proposed in draft-quinn-nsh-00. =20 On 6/20/13 7:23 AM, "Adrian Farrel" wrote: >Hi, > >I know this cuts into use cases, but I think it will turn out to be >important >for clarity and scope. > >There is a risk that "we all know what we mean", but that no-one clearly >articulates what "network service chaining" is. (It is a little like NFV >where >the definition also seems to be hard to pin down.) > >I don't want to rat-hole too much on definitions, but it would help if we >can >draw some lines. When is "service chaining" not "network service >chaining"? > >Thanks, >Adrian > >_______________________________________________ >nsc mailing list >nsc@ietf.org >https://www.ietf.org/mailman/listinfo/nsc --_002_68B171751455884590F8E38E96416F363F53F9F7xmbrcdx01ciscoc_ Content-Type: application/xml; name="default.xml" Content-Description: default.xml Content-Disposition: attachment; filename="default.xml"; size=3222; creation-date="Thu, 20 Jun 2013 14:59:34 GMT"; modification-date="Thu, 20 Jun 2013 14:59:34 GMT" Content-ID: Content-Transfer-Encoding: base64 UEsDBBQABgAIAAAAIQCb6HBP/AAAABwCAAATAAAAW0NvbnRlbnRfVHlwZXNdLnhtbKyRy2rDMBBF 94X+g9C22HK6KKXYzqKPXR+L9AMGeWyL2CMhTULy9x07LpQSAoVuBNLMvffMqFwfxkHtMSbnqdKr vNAKyfrGUVfpz81Ldq9VYqAGBk9Y6SMmva6vr8rNMWBSoqZU6Z45PBiTbI8jpNwHJKm0Po7Aco2d CWC30KG5LYo7Yz0xEmc8eei6fMIWdgOr54M8n0hErtXjqW+KqjSEMDgLLKBmqpqzuohDuiDcU/OL LlvIclHO5ql3Id0sCe+ymugaVB8Q+Q1G4TAsQ+LP8xVIRov5ZeYz0b5tncXG290o68hn48XsTwCr /4n+zjTz39ZfAAAA//8DAFBLAwQUAAYACAAAACEApdan58AAAAA2AQAACwAAAF9yZWxzLy5yZWxz hI/PasMwDIfvhb2D0X1R0sMYJXYvpZBDL6N9AOEof2giG9sb69tPxwYKuwiEpO/3qT3+rov54ZTn IBaaqgbD4kM/y2jhdj2/f4LJhaSnJQhbeHCGo3vbtV+8UNGjPM0xG6VItjCVEg+I2U+8Uq5CZNHJ ENJKRds0YiR/p5FxX9cfmJ4Z4DZM0/UWUtc3YK6PqMn/s8MwzJ5PwX+vLOVFBG43lExp5GKhqC/j U72QqGWq1B7Qtbj51v0BAAD//wMAUEsDBBQABgAIAAAAIQBreZYWgwAAAIoAAAAcAAAAdGhlbWUv dGhlbWUvdGhlbWVNYW5hZ2VyLnhtbAzMTQrDIBBA4X2hd5DZN2O7KEVissuuu/YAQ5waQceg0p/b 1+XjgzfO3xTVm0sNWSycBw2KZc0uiLfwfCynG6jaSBzFLGzhxxXm6XgYybSNE99JyHNRfSPVkIWt td0g1rUr1SHvLN1euSRqPYtHV+jT9yniResrJgoCOP0BAAD//wMAUEsDBBQABgAIAAAAIQAhWqKE IQcAANsdAAAWAAAAdGhlbWUvdGhlbWUvdGhlbWUxLnhtbOxZT28bRRS/I/EdRnsvsRMnTaI6VezY DbRpo9gt6nG8O/ZOM7uzmhkn8Q21RyQkREEcqMSNAwIqtRKX8mkCRVCkfgXezOyud+Jxk5QAFTSH 1jv7e2/e+70/82evXD1KGDogQlKeNoP6e7UAkTTkEU1HzeB2v3tpNUBS4TTCjKekGUyIDK5uvPvO FbyuYpIQBPKpXMfNIFYqW19YkCEMY/kez0gK74ZcJFjBoxgtRAIfgt6ELSzWaisLCaZpgFKcgNpb wyENCeprlcFGobzD4DFVUg+ETPS0auJIGGy0X9cIOZFtJtABZs0A5on4YZ8cqQAxLBW8aAY18xcs bFxZwOu5EFNzZCtyXfOXy+UC0f6imVOMBuWk9W5j7fJWqd8AmJrFdTqddqde6jMAHIbgqbWlqrPR Xa23Cp0VkP05q7tdW641XHxF/9KMzWutVmt5LbfFKjUg+7Mxg1+trTQ2Fx28AVn88gy+0dpst1cc vAFZ/MoMvnt5baXh4g0oZjTdn0HrgHa7ufYSMuRs2wtfBfhqLYdPUZANZXbpKYY8VfNyLcH3uOgC QAMZVjRFapKRIQ4hi9uY0YGgegK8TnDljR0K5cyQngvJUNBMNYMPMgwVMdX38tl3L589Qcf3nx7f //H4wYPj+z9YRY7UNk5HVakX33z6x6OP0O9Pvn7x8HM/Xlbxv3z/8c8/feYHQvlMzXn+xeNfnz5+ /uUnv3370APfFHhQhfdpQiS6SQ7RHk/AMcOKazkZiPNJ9GNMqxKb6UjiFOtZPPo7KnbQNyeYYQ+u RVwG7whoHz7gtfE9x+BeLMYqj7fj2fU4cYA7nLMWF14Wruu5KjT3x+nIP7kYV3F7GB/45m7j1Ilv Z5xB36Q+le2YOGbuMpwqPCIpUUi/4/uEePi6S6nD6w4NBZd8qNBdilqYeinp04GTTVOhbZpAXCY+ AyHeDjc7d1CLM5/XW+TARUJVYOYxvk+YQ+M1PFY48ans44RVCb+BVewzsjcRYRXXkQoiPSKMo05E pPTJ3BLgbyXo16F1+MO+wyaJixSK7vt03sCcV5FbfL8d4yTzYXs0javY9+U+pChGu1z54DvcrRD9 DHHA6dxw36HECffp3eA2HTkmTRNEvxkLTyyvEe7kb2/ChpiYVgNN3enVCU1f1bgT6Nu54xfXuKFV Pv/qkcfuN7VlbwIJvprZPtGo5+FOtuc2FxF987vzFh6nuwQKYnaJetuc3zbn4D/fnOfV88W35GkX hgatt0x2o2223cncXfeQMtZTE0ZuSLPxlrD2RF0Y1HLmxEnKU1gWw09dyTCBgxsJbGSQ4OpDquJe jDPYtNcDrWQkc9UjiTIu4bBohr26NR42/soeNZf1IcR2DonVDo/s8JIeLs4apRpj1cgcaIuJlrSC s062dDlXCr69zmR1bdSZZ6sb00xTdGYrXdYUm0M5UF66BoMlm7CpQbAVApZX4Myvp4bDDmYk0rzb GBVhMVH4e0KUe20diXFEbIic4QqbdRO7IoVm/NPu2Rw5H5sla0Da6UaYtJifP2ckuVAwJRkET1YT S6u1xVJ02AzWlheXAxTirBkM4ZgLP5MMgib1NhCzEdwVhUrYrD21Fk2RTj1e82dVHW4u5hSMU8aZ kGoLy9jG0LzKQ8VSPZO1f3G5oZPtYhzwNJOzWbG0Cinyr1kBoXZDS4ZDEqpqsCsjmjv7mHdCPlZE 9OLoEA3YWOxhCD9wqv2JqITbClPQ+gGu1jTb5pXbW/NOU73QMjg7jlkW47xb6quZouIs3PST0gbz VDEPfPPabpw7vyu64i/KlWoa/89c0csBXB4sRToCIdzsCox0pTQDLlTMoQtlMQ27AtZ90zsgW+B6 Fl4D+XC/bP4X5ED/b2vO6jBlDWdAtUdHSFBYTlQsCNmFtmSy7xRl9XzpsSpZrshkVMVcmVmzB+SA sL7ugSu6BwcohlQ33SRvAwZ3Mv/c57yCBiO9R6nWm9PJyqXT1sA/vXGxxQxOndhL6Pwt+C9NLFf3 6epn5Y14sUZWHdEvprukRlEVzuK3tpZP9ZomnGUBrqy1tmPNeLy4XBgHUZz1GAbL/UwGV0BI/wPr HxUhsx8r9ILa53vQWxF8e7D8IcjqS7qrQQbpBml/DWDfYwdtMmlVltp856NZKxbrC96olvOeIFtb dpZ4n5PschPlTufU4kWSnTPscG3H5lINkT1ZojA0LM4hJjDmK1f1QxQf3INAb8GV/5jZT1MygydT B9muMNk14NEk/8mkXXBt1ukzjEaydI8MEY2OivNHyYQtIft5pNgiG7QW04lWCi75Dg2uYI7Xona1 LIUXTxcuJczM0LJLYXOX5lMAH8fyxq2PdoC3TdZ6rYurYIqlf4WyMxjvp8x78jkrZfag+MpAvQZl 6ujVlOVMAXmziQefNwWGo1fP9F9YdGymm5Td+BMAAP//AwBQSwMEFAAGAAgAAAAhAA3RkJ+2AAAA GwEAACcAAAB0aGVtZS90aGVtZS9fcmVscy90aGVtZU1hbmFnZXIueG1sLnJlbHOEj00KwjAUhPeC dwhvb9O6EJEm3YjQrdQDhOQ1DTY/JFHs7Q2uLAguh2G+mWm7l53JE2My3jFoqhoIOumVcZrBbbjs jkBSFk6J2TtksGCCjm837RVnkUsoTSYkUiguMZhyDidKk5zQilT5gK44o49W5CKjpkHIu9BI93V9 oPGbAXzFJL1iEHvVABmWUJr/s/04GolnLx8WXf5RQXPZhQUoosbM4CObqkwEylu6usTfAAAA//8D AFBLAQItABQABgAIAAAAIQCb6HBP/AAAABwCAAATAAAAAAAAAAAAAAAAAAAAAABbQ29udGVudF9U eXBlc10ueG1sUEsBAi0AFAAGAAgAAAAhAKXWp+fAAAAANgEAAAsAAAAAAAAAAAAAAAAALQEAAF9y ZWxzLy5yZWxzUEsBAi0AFAAGAAgAAAAhAGt5lhaDAAAAigAAABwAAAAAAAAAAAAAAAAAFgIAAHRo ZW1lL3RoZW1lL3RoZW1lTWFuYWdlci54bWxQSwECLQAUAAYACAAAACEAIVqihCEHAADbHQAAFgAA AAAAAAAAAAAAAADTAgAAdGhlbWUvdGhlbWUvdGhlbWUxLnhtbFBLAQItABQABgAIAAAAIQAN0ZCf tgAAABsBAAAnAAAAAAAAAAAAAAAAACgKAAB0aGVtZS90aGVtZS9fcmVscy90aGVtZU1hbmFnZXIu eG1sLnJlbHNQSwUGAAAAAAUABQBdAQAAIwsAAAAA --_002_68B171751455884590F8E38E96416F363F53F9F7xmbrcdx01ciscoc_-- From narten@us.ibm.com Thu Jun 20 08:37:48 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 15C7B21F9D6E for ; Thu, 20 Jun 2013 08:37:46 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -110.599 X-Spam-Level: X-Spam-Status: No, score=-110.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8, USER_IN_WHITELIST=-100] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id hrfeOYAzCIOw for ; Thu, 20 Jun 2013 08:37:39 -0700 (PDT) Received: from e38.co.us.ibm.com (e38.co.us.ibm.com [32.97.110.159]) by ietfa.amsl.com (Postfix) with ESMTP id DE02621F9D7A for ; Thu, 20 Jun 2013 08:37:32 -0700 (PDT) Received: from /spool/local by e38.co.us.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Thu, 20 Jun 2013 09:27:17 -0600 Received: from d03dlp03.boulder.ibm.com (9.17.202.179) by e38.co.us.ibm.com (192.168.1.138) with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted; Thu, 20 Jun 2013 09:27:14 -0600 Received: from d03relay03.boulder.ibm.com (d03relay03.boulder.ibm.com [9.17.195.228]) by d03dlp03.boulder.ibm.com (Postfix) with ESMTP id DED0419D8048 for ; Thu, 20 Jun 2013 09:27:03 -0600 (MDT) Received: from d03av04.boulder.ibm.com (d03av04.boulder.ibm.com [9.17.195.170]) by d03relay03.boulder.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id r5KFRA6a106588 for ; Thu, 20 Jun 2013 09:27:11 -0600 Received: from d03av04.boulder.ibm.com (loopback [127.0.0.1]) by d03av04.boulder.ibm.com (8.14.4/8.13.1/NCO v10.0 AVout) with ESMTP id r5KFPlCI015021 for ; Thu, 20 Jun 2013 09:25:47 -0600 Received: from cichlid.raleigh.ibm.com (sig-9-76-159-14.mts.ibm.com [9.76.159.14]) by d03av04.boulder.ibm.com (8.14.4/8.13.1/NCO v10.0 AVin) with ESMTP id r5KFPklO014917; Thu, 20 Jun 2013 09:25:46 -0600 Received: from cichlid.raleigh.ibm.com (localhost [127.0.0.1]) by cichlid.raleigh.ibm.com (8.14.4/8.12.5) with ESMTP id r5KFKf2k005603; Thu, 20 Jun 2013 11:20:41 -0400 Message-Id: <201306201520.r5KFKf2k005603@cichlid.raleigh.ibm.com> To: "Jim Guichard (jguichar)" In-reply-to: <68B171751455884590F8E38E96416F363F53F9F7@xmb-rcd-x01.cisco.com> References: <68B171751455884590F8E38E96416F363F53F9F7@xmb-rcd-x01.cisco.com> Comments: In-reply-to "Jim Guichard (jguichar)" message dated "Thu, 20 Jun 2013 14:59:34 -0000." Date: Thu, 20 Jun 2013 11:20:41 -0400 From: Thomas Narten X-TM-AS-MML: No X-Content-Scanned: Fidelis XPS MAILER x-cbid: 13062015-5518-0000-0000-00000FC1C2BD Cc: "nsc@ietf.org" , "adrian@olddog.co.uk" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 15:37:48 -0000 Personally, I think the term "service" is OK. Yes, it's somewhat overloaded, but almost any term we use will be. But we do need to define what we mean by the term That said ... "Jim Guichard (jguichar)" writes: > From my perspective "service" is an overloaded term, so we attempted to > bound the concept of service. Depending on perspective a service could be > an individual vertical service such as a financial or sales application, a > transit service such as network security or IP connectivity, a service > providing compute or storage resources, a network-based service such as a > firewall, a particular piece of content such as a video stream, or some > combination of all of these. This is the general notion of a service, but I think its too broad for NSC. Don't we really only care about services where the actual source & destination of the traffic are: 1) unaware that the service is being applied and 2) do not address that service explicitely? I.e., what we usually think of as middleboxes? Because end points don't address them, normal packet forwarding doesn't ensure traffic always goes through the "service". And NSC is all about being able to describe the existance of such services and having mechanisms that ensure all traffic (as defined by flows/policy) flows through the services that are required to be applied. So, NSC is needed to force traffic flows to properly traverse middlebox "services". But NSC is not needed to have traffic delivered to say, a storage controller. Right? Thomas From paulq@cisco.com Thu Jun 20 09:01:34 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0D38A21F9D4D for ; Thu, 20 Jun 2013 09:01:34 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.499 X-Spam-Level: X-Spam-Status: No, score=-10.499 tagged_above=-999 required=5 tests=[AWL=0.100, BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id x6uNmUg8XH0c for ; Thu, 20 Jun 2013 09:01:29 -0700 (PDT) Received: from mtv-iport-4.cisco.com (mtv-iport-4.cisco.com [173.36.130.15]) by ietfa.amsl.com (Postfix) with ESMTP id 9C71021F99B0 for ; Thu, 20 Jun 2013 09:01:29 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=3637; q=dns/txt; s=iport; t=1371744089; x=1372953689; h=mime-version:subject:from:in-reply-to:date:cc: content-transfer-encoding:message-id:references:to; bh=S86T5CqbWLmRRL1ZbovimD3Mxg6qPuMuTMPNZoRf4zU=; b=FqRKHvO725fPgk545bq2H9G0n/lkRmMKePmdBphtlRHyUdjBNxwzSRkd dtzD4bkYamnw1xFtlH7PoZqTEK0kZKlTRcOrKjnE52ufC0NwwdAZpkjHG IwzBNSe4Ut+I86CjI0nJFUulMBiIkEMnVQ/XifTpu87NKgIrLz+YvXbE8 g=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AgQFANcmw1GrRDoI/2dsb2JhbABbgwkxv398FnSCIwEBAQMBAQEBawsFCwsOCi4nMAYTiAgFDbwXBI8WMweDAGEDiSGOIpFEgy8c X-IronPort-AV: E=Sophos;i="4.87,906,1363132800"; d="scan'208";a="83929129" Received: from mtv-core-3.cisco.com ([171.68.58.8]) by mtv-iport-4.cisco.com with ESMTP; 20 Jun 2013 16:01:28 +0000 Received: from sjc-paquinn-8713.cisco.com (sjc-paquinn-8713.cisco.com [10.19.172.244]) by mtv-core-3.cisco.com (8.14.5/8.14.5) with ESMTP id r5KG1RB4006906 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO); Thu, 20 Jun 2013 16:01:27 GMT Content-Type: text/plain; charset=windows-1252 Mime-Version: 1.0 (Mac OS X Mail 6.5 \(1508\)) From: Paul Quinn In-Reply-To: Date: Thu, 20 Jun 2013 09:01:27 -0700 Content-Transfer-Encoding: quoted-printable Message-Id: <68B30924-770A-48D5-908F-31918D25E4BB@cisco.com> References: To: Ken Gray X-Mailer: Apple Mail (2.1508) Cc: "nsc@ietf.org" Subject: Re: [nsc] draft-quinn-nsh questions X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 16:01:34 -0000 Hi Ken, Thanks for the great questions. Please see my replies inline. Paul On Jun 19, 2013, at 7:01 PM, Ken Gray wrote: > Okay, out here in the open then ... >=20 > For the header doc, I had two questions ... >=20 > 1) The requirement for the context headers to be present if > they're unused. Do you have a feel for what the reaction to the 128 = bit > overhead is going to be? If the context could be collapsed when > unused, it could cushion the blow of the overhead, even though it may > complicate header offset setting in machine microcode - I mean, it's a > tradeoff. Interested in your thoughts there =85 >=20 PQ> We discussed the trade-off and decided the simplicity of a single = fixed sized header is worth the potential extra overhead. Given that most of the initial = service chaining use cases are=20 focused in the data center, with planned NSH ASIC platforms (forwarding = at multiples of=20 10/40/100G) the overhead is relatively minimal.=20 >=20 > 2) I love the idea of the proxy. It applies to already > existing virtual or non-virtual appliances so allows us to suck in = vApps > that were designed for other chaining schemes (and for a while this is > going to be the wild-wild west), even proprietary ones. > So, is the vision that this be a user space agent that is just = "canned" > as far as it's processing goes (like a utility function) - that is, = it's > a vendor-independent piece of software (potentially open-sourced) > and/or a similar thing but with hooks to vendor specific overlay = mapping > software to help with the encap/decap? PQ> A "canned" or open source piece of software is precisely what we = have in mind. =20 This "agent" can reside in a physical switch, a vswitch, or some other = proxy node and=20 participate in the overlay. The only hooks needed are to present to = the chaining control system the location of the services in the dataplane. > I mean, do I have it wrong or > would the expectation be that this manipulates no metadata as there is > no real context sharing with non-compliant apps - or am I missing > something and it is more complex/customizable? >=20 >=20 PQ> Metadata manipulation is interesting in this case. The proxy must = be able to=20 preserve the existing metadata. As a base function, a proxy will map = the path ID to a=20 locally significant identifier that the service functions can = understand. However, an=20 interesting evolution to that proxy functions is to offer configurable = mapping of more=20 than just the path ID. So perhaps, in some deployments, metadata (or a = combination of=20 fields) is mapped to locally significant identifier(s). > On 6/19/13 5:44 PM, "Paul Quinn" wrote: >=20 >> Fellow NSCers, >>=20 >> In preparation for Berlin, I'd like to kick of some discussion on the >> list: >>=20 >> 1. How do you feel about the topic of service chaining in general? >> There's a lot of discussion in the industry and amongst operators = about >> it, but on this list, what is the general thought about how service >> chaining will impact networks? >>=20 >> 2. I'm hoping folks have started to read the documents posted with = the >> BoF request. Please send you comments to the list. In addition to = those >> documents, what else do you expect to see addressed by this effort? >>=20 >> Thanks >> Paul >> _______________________________________________ >> nsc mailing list >> nsc@ietf.org >> https://www.ietf.org/mailman/listinfo/nsc >>=20 >>=20 >=20 >=20 >=20 From bensons@queuefull.net Thu Jun 20 09:37:58 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 13B0C21F9E3E for ; Thu, 20 Jun 2013 09:37:58 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.599 X-Spam-Level: X-Spam-Status: No, score=-2.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id L4adKdRAl2EH for ; Thu, 20 Jun 2013 09:37:57 -0700 (PDT) Received: from mail-ve0-x232.google.com (mail-ve0-x232.google.com [IPv6:2607:f8b0:400c:c01::232]) by ietfa.amsl.com (Postfix) with ESMTP id 264A621F9E32 for ; Thu, 20 Jun 2013 09:37:57 -0700 (PDT) Received: by mail-ve0-f178.google.com with SMTP id pb11so5218757veb.23 for ; Thu, 20 Jun 2013 09:37:56 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=message-id:date:from:user-agent:mime-version:to:cc:subject :references:in-reply-to:content-type:content-transfer-encoding :x-gm-message-state; bh=/pOz1ydCUdoLrWL1apTW7pDphX5sqskghQ+pKjNNYAo=; b=NxXyzqtXCedNwKgZDfIy5gggZc9VlnbzFXQ6FSYlsTtCImTKhaKaABDPpmXm3jtaTb rfnSun5+XD5lxeF489Vd45/3OPsl7TmhvzKfURzPDQytdMhpJcCoMz6c9wd/zW+clom3 Co7+hF1PNXcElazrjvJDhBmT5raMwyoQUcAypVd6yil86okrk4SQdQkTJjxWkFX+vnQX fnHT6nTYFjGCPxyqdt0jVfbuP4fxp5NcMhiU0XCuBvmcqwNYETSj0ke1yKKgSvLaO6to a3mFHid4e8/1uf15zK5RbjzpUvcME0TIPTxJr/WqiS6+xzKGJBDDoEsxb+1AzC4b66G2 zoTA== X-Received: by 10.52.170.11 with SMTP id ai11mr2956138vdc.97.1371746276600; Thu, 20 Jun 2013 09:37:56 -0700 (PDT) Received: from fallout.local (68-115-154-254.static.hckr.nc.charter.com. [68.115.154.254]) by mx.google.com with ESMTPSA id wl4sm1176062vdb.3.2013.06.20.09.37.55 for (version=TLSv1 cipher=ECDHE-RSA-RC4-SHA bits=128/128); Thu, 20 Jun 2013 09:37:55 -0700 (PDT) Message-ID: <51C32FE2.90108@queuefull.net> Date: Thu, 20 Jun 2013 12:37:54 -0400 From: Benson Schliesser User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.6; rv:17.0) Gecko/20130509 Thunderbird/17.0.6 MIME-Version: 1.0 To: adrian@olddog.co.uk References: <07c001ce6da8$904f9130$b0eeb390$@olddog.co.uk> <94C682931C08B048B7A8645303FDC9F36EDB259223@PUEXCB1B.nanterre.francetelecom.fr> <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> In-Reply-To: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Gm-Message-State: ALoCoQkUwcP9Ovdn0kHAyv2qwcfxSvUfgHtCKJ0h5YwmZ9X0ssetXYblF2ehT5Xu/GFytJOLIMeJ Cc: nsc@ietf.org Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 16:37:58 -0000 Hi, Adrian. On 2013-06-20 10:26 AM, Adrian Farrel wrote: > Something I pick out of this is that all the example functions are > packet-oriented functions (or possibly flow-oriented) and not > application layer services. I think this is significant and not > accidental :-) Can we capture that idea in a definition? We do need to agree on what a "network service" is, as distinguished from non-network services. I suspect that some people intuitively lump together many things as network services, that should really be further divided into "Endpoint Network Services" and "Inline Network Services". (Hopefully somebody can come up with better category names; I'm just trying to make the general point.) For example, DNS and DHCP could be examples of Endpoint Network Services, while Firewall or NAT could be examples of Inline Network Services. The key difference, as both you and Thomas Narten have pointed out, is whether packets/flows are terminating (at the Endpoint) or being forwarded (through the Inline service). Since we're focused on Service Chaining, this group needs to investigate how packets/flows are routed through Inline Network Services. It seems pretty obvious that we already have the means to get packets to the correct endpoint, and so I assume Endpoint Network Services are out of scope when we talk about "Network Services" here. (But, of course, somebody could convince me otherwise.) I'm hesitant to categorize based on whether something works at the "application layer", because I can imagine Inline Network Services that do so yet still need Service Chaining routing etc. Cheers, -Benson From bensons@queuefull.net Thu Jun 20 09:39:52 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 5768621F9E9C for ; Thu, 20 Jun 2013 09:39:52 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.599 X-Spam-Level: X-Spam-Status: No, score=-2.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id rsbFqTHnG+zD for ; Thu, 20 Jun 2013 09:39:51 -0700 (PDT) Received: from mail-vb0-x236.google.com (mail-vb0-x236.google.com [IPv6:2607:f8b0:400c:c02::236]) by ietfa.amsl.com (Postfix) with ESMTP id 64F2621F9EA0 for ; Thu, 20 Jun 2013 09:39:45 -0700 (PDT) Received: by mail-vb0-f54.google.com with SMTP id q12so4885799vbe.13 for ; Thu, 20 Jun 2013 09:39:44 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=message-id:date:from:user-agent:mime-version:to:cc:subject :references:in-reply-to:content-type:content-transfer-encoding :x-gm-message-state; bh=UP2nS6MtOYVkvqEw9WJORf+VkrhmFGL24h0lXkZl/vU=; b=mSmEaTd/9hlKrBgRdZ+siONLl4o1UNHkxT1XS5Mm89lXrI+CBuYGYwVVkOAXBcNQQe TsERyJIqVtoJZHihLlC1kl7hZGoqqDffFCe66Rw0PLLyItRp0m5q5WmmkDs0FkkhtEEV N8b+1h0utBXWKG38e6RF3ZOza4Lv694YFxrnQw72+O5pxvj+nubsUOLlzxK/Pa8EoXT0 aEQscWI4fB0Pb08NxCZgqvVsChVkIJpsfg8SlTo3EYTrQj5EDCP5yXinVx8uf5u2+SZ4 uMgRF5QbfOx9fHHfWaR8K8s7414VH6NMofUQ/iS6RPH0k0ksCr3Jq6yjjobKoJCLRAFP +/Fw== X-Received: by 10.58.255.229 with SMTP id at5mr3522741ved.44.1371746384709; Thu, 20 Jun 2013 09:39:44 -0700 (PDT) Received: from fallout.local (68-115-154-254.static.hckr.nc.charter.com. [68.115.154.254]) by mx.google.com with ESMTPSA id at19sm1130257vdc.10.2013.06.20.09.39.43 for (version=TLSv1 cipher=ECDHE-RSA-RC4-SHA bits=128/128); Thu, 20 Jun 2013 09:39:43 -0700 (PDT) Message-ID: <51C3304E.9070004@queuefull.net> Date: Thu, 20 Jun 2013 12:39:42 -0400 From: Benson Schliesser User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.6; rv:17.0) Gecko/20130509 Thunderbird/17.0.6 MIME-Version: 1.0 To: Thomas Narten References: <68B171751455884590F8E38E96416F363F53F9F7@xmb-rcd-x01.cisco.com> <201306201520.r5KFKf2k005603@cichlid.raleigh.ibm.com> In-Reply-To: <201306201520.r5KFKf2k005603@cichlid.raleigh.ibm.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Gm-Message-State: ALoCoQl1lJaJvX889JXIB1uO7mwN0PsE3r8cty1kXA45nl1g/lbDjuaTA+zZdn9YnthDDU6RQ9Hl Cc: "nsc@ietf.org" , "Jim Guichard \(jguichar\)" , "adrian@olddog.co.uk" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 16:39:52 -0000 On 2013-06-20 11:20 AM, Thomas Narten wrote: > Personally, I think the term "service" is OK. Yes, it's somewhat > overloaded, but almost any term we use will be. But we do need to > define what we mean by the term I don't think it really matters if we call it a "service" or "function" - either term is broad enough. But personally I like the term "service". In my mind at least, there could be a subtle distinction between a "service" and a "function" in that the word "function" could describe a lower-level capability, whereas the word "service" describes something more complete. E.g. multiple functions might come together to provide a service (implying that "functions" are internal aspects whereas "services" are atomic things) But that's just my leaning; I don't think I can defend this as _the_ correct definition. And frankly, this group may choose to work at a level of granularity that reaches down to the "function" as I've described it. Cheers, -Benson From paulq@cisco.com Thu Jun 20 09:42:28 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7987621F9E53 for ; Thu, 20 Jun 2013 09:42:28 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.524 X-Spam-Level: X-Spam-Status: No, score=-10.524 tagged_above=-999 required=5 tests=[AWL=0.075, BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id x4EGLIooPwSg for ; Thu, 20 Jun 2013 09:42:24 -0700 (PDT) Received: from mtv-iport-4.cisco.com (mtv-iport-4.cisco.com [173.36.130.15]) by ietfa.amsl.com (Postfix) with ESMTP id 0F6F321F9E42 for ; Thu, 20 Jun 2013 09:42:24 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=4161; q=dns/txt; s=iport; t=1371746544; x=1372956144; h=mime-version:subject:from:in-reply-to:date:cc: content-transfer-encoding:message-id:references:to; bh=StC7wOSqjYMfIFPQyKCjhEeoNYwfz3bYBbyExka2Qe8=; b=g9KucT3ALpFSJzsKLzt7okz8ObOdPmUem/KZQ3xSaaDbmbZPE+yO/n0R XcORm6PFYMdSCzvVmBdSwBgcm2CmJZ8EoRivSOvtWmAvqWy603W9WgnJ1 1gQ0CVhBckwppeVeP0qyf2g5oON15BtC6kBuj9tZVyf8unGY9KWWbB5dP 0=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AgQFADsww1GrRDoI/2dsb2JhbABbgwkxv398FnSCIwEBAQMBAQEBNzQLBQsLEgYuJyIOBhOICAUNvBYEjgeBDzMHgwBhA4khjiKRRIMvHIE1 X-IronPort-AV: E=Sophos;i="4.87,906,1363132800"; d="scan'208";a="83931105" Received: from mtv-core-3.cisco.com ([171.68.58.8]) by mtv-iport-4.cisco.com with ESMTP; 20 Jun 2013 16:42:19 +0000 Received: from sjc-paquinn-8713.cisco.com (sjc-paquinn-8713.cisco.com [10.19.172.244]) by mtv-core-3.cisco.com (8.14.5/8.14.5) with ESMTP id r5KGgIWe005183 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO); Thu, 20 Jun 2013 16:42:18 GMT Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 (Mac OS X Mail 6.5 \(1508\)) From: Paul Quinn In-Reply-To: <51C2650B.8040502@joelhalpern.com> Date: Thu, 20 Jun 2013 09:42:18 -0700 Content-Transfer-Encoding: quoted-printable Message-Id: <783EE707-F570-482B-A7AA-6973AC4C0BCF@cisco.com> References: <51C2650B.8040502@joelhalpern.com> To: "Joel M. Halpern" X-Mailer: Apple Mail (2.1508) Cc: "nsc@ietf.org" Subject: Re: [nsc] draft-quinn-nsh questions X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 16:42:28 -0000 Hi Joel, Thanks for your email. First off, it's great to hear that overall, many = of us agree that=20 service chaining is an worthy topic :) As you probably saw, we submitted two documents with the BoF proposal. = Our goal with those documents is twofold: i) describe a problem, that we can discuss = as a group at the BoF, and eventually a WG, ii) reflect our implementation that addresses = some of the problems outlined in the problem statement. Given how "hot" service chaining is, we opted to be pragmatic and to = quickly start=20 offering guidance for implementors, based on work we are doing;=20 starting down the interoperability path seemed reasonable. Of course, in Berlin, we'll get together as a group, and decide what = other drafts are=20 needed. Thanks Paul On Jun 19, 2013, at 7:12 PM, Joel M. Halpern = wrote: > Service Chaining is clearly a topic of interest to man of us = individually and many of our companies either as vendors, service = providers, or users of services. That's the easy part. >=20 > I find myself concerned seeing a number of proposals for specific = protocol mechanism when there is nothing resembling agreement on the = needs, scales, or architecture for the problem. The number of sets of = non-matching assumptions out there is large enough that I am having = difficulty getting my head around what a working group would do. = Defining protocol seems an odd place to start. >=20 > Yours, > Joel >=20 > On 6/19/2013 10:01 PM, Ken Gray wrote: >> Okay, out here in the open then ... >>=20 >> For the header doc, I had two questions ... >>=20 >> 1) The requirement for the context headers to be present if >> they're unused. Do you have a feel for what the reaction to the 128 = bit >> overhead is going to be? If the context could be collapsed when >> unused, it could cushion the blow of the overhead, even though it may >> complicate header offset setting in machine microcode - I mean, it's = a >> tradeoff. Interested in your thoughts there ... >>=20 >>=20 >> 2) I love the idea of the proxy. It applies to already >> existing virtual or non-virtual appliances so allows us to suck in = vApps >> that were designed for other chaining schemes (and for a while this = is >> going to be the wild-wild west), even proprietary ones. >> So, is the vision that this be a user space agent that is just = "canned" >> as far as it's processing goes (like a utility function) - that is, = it's >> a vendor-independent piece of software (potentially open-sourced) >> and/or a similar thing but with hooks to vendor specific overlay = mapping >> software to help with the encap/decap? I mean, do I have it wrong = or >> would the expectation be that this manipulates no metadata as there = is >> no real context sharing with non-compliant apps - or am I missing >> something and it is more complex/customizable? >>=20 >>=20 >> On 6/19/13 5:44 PM, "Paul Quinn" wrote: >>=20 >>> Fellow NSCers, >>>=20 >>> In preparation for Berlin, I'd like to kick of some discussion on = the >>> list: >>>=20 >>> 1. How do you feel about the topic of service chaining in general? >>> There's a lot of discussion in the industry and amongst operators = about >>> it, but on this list, what is the general thought about how service >>> chaining will impact networks? >>>=20 >>> 2. I'm hoping folks have started to read the documents posted with = the >>> BoF request. Please send you comments to the list. In addition to = those >>> documents, what else do you expect to see addressed by this effort? >>>=20 >>> Thanks >>> Paul >>> _______________________________________________ >>> nsc mailing list >>> nsc@ietf.org >>> https://www.ietf.org/mailman/listinfo/nsc >>>=20 >>>=20 >>=20 >>=20 >>=20 >> _______________________________________________ >> nsc mailing list >> nsc@ietf.org >> https://www.ietf.org/mailman/listinfo/nsc >>=20 > _______________________________________________ > nsc mailing list > nsc@ietf.org > https://www.ietf.org/mailman/listinfo/nsc From jguichar@cisco.com Thu Jun 20 09:50:56 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 8D8AD21F9984 for ; Thu, 20 Jun 2013 09:50:55 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.599 X-Spam-Level: X-Spam-Status: No, score=-10.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id kuzJOgzvKLLb for ; Thu, 20 Jun 2013 09:50:49 -0700 (PDT) Received: from rcdn-iport-3.cisco.com (rcdn-iport-3.cisco.com [173.37.86.74]) by ietfa.amsl.com (Postfix) with ESMTP id 28E0921F9D64 for ; Thu, 20 Jun 2013 09:50:49 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=1176; q=dns/txt; s=iport; t=1371747049; x=1372956649; h=from:to:cc:subject:date:message-id:in-reply-to: content-id:content-transfer-encoding:mime-version; bh=qISm4gUD31o/CRPI51s/P1CZQ4iCpjNAMTe6cdtoNu0=; b=lgf4A6l/cOo3kClTHjoLC56BL+Z80r/Rnzz0TQu5wlNd2BcDYXAxu6jQ q/vFZCgZLBuOjpHNT6W8XZewaCtyg8RvT3VjjqEoy4ST3nxv+RwpC8T5D +sXCFVylXppqEjT3kdlPHwoptb9oEAvOxaFaRKVcd3UxmozYdCPhuJ4ZY w=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AgcFAJAyw1GtJXHA/2dsb2JhbABbgwl6viyBCnwWdIIjAQEBBDo/EgEIGAoUQiUBAQQBDQUIiAa8J48YMQeDAGEDoXKHFYMPgig X-IronPort-AV: E=Sophos;i="4.87,906,1363132800"; d="scan'208";a="225428143" Received: from rcdn-core2-5.cisco.com ([173.37.113.192]) by rcdn-iport-3.cisco.com with ESMTP; 20 Jun 2013 16:50:24 +0000 Received: from xhc-aln-x14.cisco.com (xhc-aln-x14.cisco.com [173.36.12.88]) by rcdn-core2-5.cisco.com (8.14.5/8.14.5) with ESMTP id r5KGoOXa026317 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Thu, 20 Jun 2013 16:50:24 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.79]) by xhc-aln-x14.cisco.com ([173.36.12.88]) with mapi id 14.02.0318.004; Thu, 20 Jun 2013 11:50:24 -0500 From: "Jim Guichard (jguichar)" To: Benson Schliesser , Thomas Narten Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAJpiAAAAke3YAAAsJ3AP//v+uA Date: Thu, 20 Jun 2013 16:50:23 +0000 Message-ID: <68B171751455884590F8E38E96416F363F53FE50@xmb-rcd-x01.cisco.com> In-Reply-To: <51C3304E.9070004@queuefull.net> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.2.3.120616 x-originating-ip: [10.98.130.212] Content-Type: text/plain; charset="us-ascii" Content-ID: <4E349810BD5A9344AB243677EA002B57@emea.cisco.com> Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Cc: "nsc@ietf.org" , "adrian@olddog.co.uk" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 16:50:57 -0000 Hi Benson, One of the reasons why I refer to them as "service functions" ;-) On 6/20/13 12:39 PM, "Benson Schliesser" wrote: > >On 2013-06-20 11:20 AM, Thomas Narten wrote: >> Personally, I think the term "service" is OK. Yes, it's somewhat >> overloaded, but almost any term we use will be. But we do need to >> define what we mean by the term > >I don't think it really matters if we call it a "service" or "function" >- either term is broad enough. But personally I like the term "service". >In my mind at least, there could be a subtle distinction between a >"service" and a "function" in that the word "function" could describe a >lower-level capability, whereas the word "service" describes something >more complete. E.g. multiple functions might come together to provide a >service (implying that "functions" are internal aspects whereas >"services" are atomic things) > >But that's just my leaning; I don't think I can defend this as _the_ >correct definition. And frankly, this group may choose to work at a >level of granularity that reaches down to the "function" as I've >described it. > >Cheers, >-Benson > From jguichar@cisco.com Thu Jun 20 10:23:24 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 33ACA21F9A75 for ; Thu, 20 Jun 2013 10:23:24 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.599 X-Spam-Level: X-Spam-Status: No, score=-10.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id E5QCdhbAkyBF for ; Thu, 20 Jun 2013 10:23:19 -0700 (PDT) Received: from rcdn-iport-5.cisco.com (rcdn-iport-5.cisco.com [173.37.86.76]) by ietfa.amsl.com (Postfix) with ESMTP id C2F4C21F949F for ; Thu, 20 Jun 2013 10:23:02 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=1669; q=dns/txt; s=iport; t=1371748983; x=1372958583; h=from:to:subject:date:message-id:in-reply-to:content-id: content-transfer-encoding:mime-version; bh=1K06dG4tVg6FjOs6JolX+gsWMKbCLBB3Ekd9wHQuFEA=; b=ADzR/qhmKS11t76WbbxuI2jOS6tI9Dd78YFzKA6A0F1dEqwnAow6T1KC 3H3Rf4iHM2oCQZlKijrDF8/gcdQc1VXCNyWJK7n+Wp2A2YNocI/79qB0U qmH6/PsH+fNSWtjQnfDgKyT/nlQ0OmSHFQQrDWNyt3KkqZyNuYYF0/wZ6 w=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AggFAFk5w1GtJV2b/2dsb2JhbABbgwkxSb4sgQp8FnSCJQEEAQEBNzQdAQgiFDcLJQEBBAESCIgGDLwcBI8YOIMAYQOhcocVgw+CKA X-IronPort-AV: E=Sophos;i="4.87,906,1363132800"; d="scan'208";a="225423387" Received: from rcdn-core-4.cisco.com ([173.37.93.155]) by rcdn-iport-5.cisco.com with ESMTP; 20 Jun 2013 17:23:02 +0000 Received: from xhc-aln-x08.cisco.com (xhc-aln-x08.cisco.com [173.36.12.82]) by rcdn-core-4.cisco.com (8.14.5/8.14.5) with ESMTP id r5KHN13I006671 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Thu, 20 Jun 2013 17:23:02 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.79]) by xhc-aln-x08.cisco.com ([173.36.12.82]) with mapi id 14.02.0318.004; Thu, 20 Jun 2013 12:23:01 -0500 From: "Jim Guichard (jguichar)" To: "adrian@olddog.co.uk" , "nsc@ietf.org" Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewABC13ID//+4ggA== Date: Thu, 20 Jun 2013 17:23:00 +0000 Message-ID: <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> In-Reply-To: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.2.3.120616 x-originating-ip: [10.98.130.212] Content-Type: text/plain; charset="us-ascii" Content-ID: <847D1689BC13254F9191FFF09533CF70@emea.cisco.com> Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 17:23:24 -0000 Hi Adrian, On 6/20/13 10:26 AM, "Adrian Farrel" wrote: >Thanks to Ken and Med for replies. > >Choosing to respond on Med's thread because he gave... > >> o NLF (Network-Located Function): refers to a function which is >> enabled in the network operated by an administrative entity. NLF >> can be for example: firewall (e.g., [RFC6092]), DPI (Deep Packet >> Inspection), LI (Lawful Intercept) module, NAT44 [RFC3022], NAT64 >> [RFC6146], DS-Lite AFTR [RFC6333], NPTv6 [RFC6296], HOST_ID >> injection, HTTP Header Enrichment function, TCP optimizer, etc. > >Something I pick out of this is that all the example functions are >packet-oriented functions (or possibly flow-oriented) and not application >layer >services. I think this is significant and not accidental :-) Jim> actually the intent was that both packet-oriented and application layer service functions are within scope. It is certainly possible to envisage service chains that incorporate both of these types of functions and we should not restrict our work to one or the other but rather let the use cases dictate. =20 >Can we capture that idea in a definition? > >I note that it is hinted at in Paul and Jim's BoF proposal text and it is >there >because I asked them why this proposal was aimed at the Routing Area not >the >Applications Area. To which the answer was "because we want to steer >packets or >flows to the place where the next service will be applied." > >Adrian > >_______________________________________________ >nsc mailing list >nsc@ietf.org >https://www.ietf.org/mailman/listinfo/nsc From paulq@cisco.com Thu Jun 20 11:04:14 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 1E88F21FA000 for ; Thu, 20 Jun 2013 11:04:14 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.539 X-Spam-Level: X-Spam-Status: No, score=-10.539 tagged_above=-999 required=5 tests=[AWL=0.060, BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id XaGu4zorfj7j for ; Thu, 20 Jun 2013 11:04:08 -0700 (PDT) Received: from mtv-iport-2.cisco.com (mtv-iport-2.cisco.com [173.36.130.13]) by ietfa.amsl.com (Postfix) with ESMTP id 50A9121F9E0D for ; Thu, 20 Jun 2013 11:04:08 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=2193; q=dns/txt; s=iport; t=1371751448; x=1372961048; h=mime-version:subject:from:in-reply-to:date:cc: content-transfer-encoding:message-id:references:to; bh=f/086mIqYd3sZiPuYP6AZB0SJIfN+nF3yAyRDx4QsmQ=; b=bfT7o2Z5duWZEtoaikbOP3I7OI18WDlOrtaNKVwfG2LzJ79W6oJven62 uG8hEqJ8FtFPsxapv1CuoN0aBlMwNdzPi8XW6tLXXhsxPQ34bP6OT/zzU CVJqYHTjMR4Zuk567+8le34ArUboOwCEVYO88Gf857ZMf9iuzsAJxTMrX Q=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AgUFAPVCw1GrRDoH/2dsb2JhbABbgwm/JoEKfRZ0giMBAQEDATorFAULCxguVwYTiAgFvC+OBoEQMweDAGEDiSGOIoovhxWDLxw X-IronPort-AV: E=Sophos;i="4.87,906,1363132800"; d="scan'208";a="84001955" Received: from mtv-core-2.cisco.com ([171.68.58.7]) by mtv-iport-2.cisco.com with ESMTP; 20 Jun 2013 18:04:06 +0000 Received: from sjc-paquinn-8713.cisco.com (sjc-paquinn-8713.cisco.com [10.19.172.244]) by mtv-core-2.cisco.com (8.14.5/8.14.5) with ESMTP id r5KI45MC009601 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO); Thu, 20 Jun 2013 18:04:05 GMT Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 (Mac OS X Mail 6.5 \(1508\)) From: Paul Quinn In-Reply-To: <51C32FE2.90108@queuefull.net> Date: Thu, 20 Jun 2013 11:04:04 -0700 Content-Transfer-Encoding: quoted-printable Message-Id: <9559ADF6-4288-47EB-BFBD-E1C236B185FE@cisco.com> References: <07c001ce6da8$904f9130$b0eeb390$@olddog.co.uk> <94C682931C08B048B7A8645303FDC9F36EDB259223@PUEXCB1B.nanterre.francetelecom.fr> <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <51C32FE2.90108@queuefull.net> To: Benson Schliesser X-Mailer: Apple Mail (2.1508) Cc: nsc@ietf.org, adrian@olddog.co.uk Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 18:04:14 -0000 Hi Benson, On Jun 20, 2013, at 9:37 AM, Benson Schliesser = wrote: > Hi, Adrian. >=20 > On 2013-06-20 10:26 AM, Adrian Farrel wrote: >> Something I pick out of this is that all the example functions are = packet-oriented functions (or possibly flow-oriented) and not = application layer services. I think this is significant and not = accidental :-) Can we capture that idea in a definition? >=20 > We do need to agree on what a "network service" is, as distinguished = from non-network services. I suspect that some people intuitively lump = together many things as network services, that should really be further = divided into "Endpoint Network Services" and "Inline Network Services". = (Hopefully somebody can come up with better category names; I'm just = trying to make the general point.) >=20 > For example, DNS and DHCP could be examples of Endpoint Network = Services, while Firewall or NAT could be examples of Inline Network = Services. The key difference, as both you and Thomas Narten have pointed = out, is whether packets/flows are terminating (at the Endpoint) or being = forwarded (through the Inline service). >=20 PQ> Although that distinction is broadly true, I'm not sure we need to = draw it out now. A chain couple be composed of terminating and inline = services and fall under the umbrella of network service chaining. =20 > Since we're focused on Service Chaining, this group needs to = investigate how packets/flows are routed through Inline Network = Services. It seems pretty obvious that we already have the means to get = packets to the correct endpoint, and so I assume Endpoint Network = Services are out of scope when we talk about "Network Services" here. = (But, of course, somebody could convince me otherwise.) >=20 PQ> I think both are in scope: a good example is the simple one of = server load balancer (terminating) and a firewall (inline). > I'm hesitant to categorize based on whether something works at the = "application layer", because I can imagine Inline Network Services that = do so yet still need Service Chaining routing etc. PQ> Agreed. Paul= From adrian@olddog.co.uk Thu Jun 20 13:46:33 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id AD3E621F9A1C for ; Thu, 20 Jun 2013 13:46:33 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.401 X-Spam-Level: X-Spam-Status: No, score=-2.401 tagged_above=-999 required=5 tests=[AWL=0.198, BAYES_00=-2.599] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id wTBT2mCPHX4l for ; Thu, 20 Jun 2013 13:46:27 -0700 (PDT) Received: from asmtp4.iomartmail.com (asmtp4.iomartmail.com [62.128.201.175]) by ietfa.amsl.com (Postfix) with ESMTP id 9A1B921F9E1F for ; Thu, 20 Jun 2013 13:46:27 -0700 (PDT) Received: from asmtp4.iomartmail.com (localhost.localdomain [127.0.0.1]) by asmtp4.iomartmail.com (8.13.8/8.13.8) with ESMTP id r5KKkOUa001811; Thu, 20 Jun 2013 21:46:24 +0100 Received: from 950129200 (dsl-sp-81-140-15-32.in-addr.broadbandscope.com [81.140.15.32]) (authenticated bits=0) by asmtp4.iomartmail.com (8.13.8/8.13.8) with ESMTP id r5KKkNU9001800 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO); Thu, 20 Jun 2013 21:46:23 +0100 From: "Adrian Farrel" To: Date: Thu, 20 Jun 2013 21:46:20 +0100 Message-ID: <08db01ce6df7$38e56330$aab02990$@olddog.co.uk> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Mailer: Microsoft Outlook 14.0 Thread-Index: Ac5t9y4GPlWSXJZKTc2hRZi75qXMZQ== Content-Language: en-gb Cc: stbryant@cisco.com Subject: [nsc] Whither NSC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list Reply-To: adrian@olddog.co.uk List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 20:46:33 -0000 Hi, The IAB and IESG met today to talk about the 15 (yes, one five) BoF proposals on the table. Of course, NSC was in the pile as well. We discussed NSC quite a bit, looking at the proposal text and the material in the drafts, and also considering the views of some of the I* who are actively involved in the ETSI NFV group at the moment. The was quite a suggestion that this is all too early for a BoF, not least because ETSI is not scheduled to come out with its results until end of July at the earliest. It was also suggested that discussing solutions work was very premature. In the end we compromised. We will hold a short (1 hour) non-WG forming BoF to discuss the service provider requirements and use cases. It is my intention (and, indeed the I* was pretty forceful on this point) that presentations should come only from service providers. The vendors' job is to listen and ask questions! I will be appointing chairs soon and then we can start to put the agenda together. Of course, the mailing list remains open and I hope the discussions continue. Thanks, Adrian From adrian@olddog.co.uk Thu Jun 20 13:52:32 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 2D9D621F9A31 for ; Thu, 20 Jun 2013 13:52:32 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.412 X-Spam-Level: X-Spam-Status: No, score=-2.412 tagged_above=-999 required=5 tests=[AWL=0.187, BAYES_00=-2.599] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 6wXpqQEoI8+7 for ; Thu, 20 Jun 2013 13:52:26 -0700 (PDT) Received: from asmtp4.iomartmail.com (asmtp4.iomartmail.com [62.128.201.175]) by ietfa.amsl.com (Postfix) with ESMTP id 7BBE521F9F2C for ; Thu, 20 Jun 2013 13:51:50 -0700 (PDT) Received: from asmtp4.iomartmail.com (localhost.localdomain [127.0.0.1]) by asmtp4.iomartmail.com (8.13.8/8.13.8) with ESMTP id r5KKpj3I004338; Thu, 20 Jun 2013 21:51:45 +0100 Received: from 950129200 (dsl-sp-81-140-15-32.in-addr.broadbandscope.com [81.140.15.32]) (authenticated bits=0) by asmtp4.iomartmail.com (8.13.8/8.13.8) with ESMTP id r5KKpiEm004323 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO); Thu, 20 Jun 2013 21:51:45 +0100 From: "Adrian Farrel" To: Date: Thu, 20 Jun 2013 21:51:42 +0100 Message-ID: <08dd01ce6df7$f8794590$e96bd0b0$@olddog.co.uk> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Mailer: Microsoft Outlook 14.0 Thread-Index: Ac5t9+4/nHb4mSBzSHKzsPQdrIYqFg== Content-Language: en-gb Cc: 'Thomas Narten' , jguichar@cisco.com, stbryant@cisco.com Subject: [nsc] Moving rapidly forward :chairs appointed X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list Reply-To: adrian@olddog.co.uk List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 20:52:32 -0000 I'm very pleased to announce that Thomas Narten and Jim Guichard have agreed to chair the NSC BoF in Berlin. Thomas brings a wealth of IETF experience, while Jim was one of the original proponents of the BoF. Please work with them to make this as successful as possible. Thanks, Adrian From jguichar@cisco.com Thu Jun 20 14:00:02 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id ECDC821E804E for ; Thu, 20 Jun 2013 14:00:01 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.599 X-Spam-Level: X-Spam-Status: No, score=-10.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id fzbs7vap0+3b for ; Thu, 20 Jun 2013 13:59:56 -0700 (PDT) Received: from rcdn-iport-3.cisco.com (rcdn-iport-3.cisco.com [173.37.86.74]) by ietfa.amsl.com (Postfix) with ESMTP id 5910D11E811B for ; Thu, 20 Jun 2013 13:59:56 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=2020; q=dns/txt; s=iport; t=1371761996; x=1372971596; h=from:to:cc:subject:date:message-id:in-reply-to: content-id:content-transfer-encoding:mime-version; bh=KJ1asfdvIoeFiw9NM6xmQB/t/o7jl5fuXMybbJZ8en8=; b=ChJLW/AmllxB5BWMAwWrE2Yw1GGC9NUoXrjujU8qwE1taxxD1aPPV4Xu iP9XfZ/k/qjyxmJ/8RQCVWI/di0ewFexmep55HJDValD+zQm0su6LzQa3 L9IOQvFEXYZVrl7OsaSYGPXby109GjGdKnj6ofBIALtfYMAx1LFiE4VCp Q=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AgcFADNsw1GtJXG+/2dsb2JhbABbgwl6viyBCn8WdIIlAQQ6PxIBCCIUQiUBAQQOBQiIBrxBjh17MQeDAGEDoXKHFYMPgWk/ X-IronPort-AV: E=Sophos;i="4.87,907,1363132800"; d="scan'208";a="225527316" Received: from rcdn-core2-3.cisco.com ([173.37.113.190]) by rcdn-iport-3.cisco.com with ESMTP; 20 Jun 2013 20:59:55 +0000 Received: from xhc-aln-x01.cisco.com (xhc-aln-x01.cisco.com [173.36.12.75]) by rcdn-core2-3.cisco.com (8.14.5/8.14.5) with ESMTP id r5KKxtdK016758 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Thu, 20 Jun 2013 20:59:55 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.79]) by xhc-aln-x01.cisco.com ([173.36.12.75]) with mapi id 14.02.0318.004; Thu, 20 Jun 2013 15:59:55 -0500 From: "Jim Guichard (jguichar)" To: Thomas Narten Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAJpiAAAAke3YAAA3brgA== Date: Thu, 20 Jun 2013 20:59:54 +0000 Message-ID: <68B171751455884590F8E38E96416F363F54043D@xmb-rcd-x01.cisco.com> In-Reply-To: <201306201520.r5KFKf2k005603@cichlid.raleigh.ibm.com> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.2.3.120616 x-originating-ip: [10.98.130.212] Content-Type: text/plain; charset="us-ascii" Content-ID: <61F0A0E894810547B99D8EE163C84170@emea.cisco.com> Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Cc: "nsc@ietf.org" , "adrian@olddog.co.uk" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Jun 2013 21:00:02 -0000 Hi Thomas, On 6/20/13 11:20 AM, "Thomas Narten" wrote: >Personally, I think the term "service" is OK. Yes, it's somewhat >overloaded, but almost any term we use will be. But we do need to >define what we mean by the term > >That said ... > >"Jim Guichard (jguichar)" writes: > >> From my perspective "service" is an overloaded term, so we attempted to >> bound the concept of service. Depending on perspective a service could >>be >> an individual vertical service such as a financial or sales >>application, a >> transit service such as network security or IP connectivity, a service >> providing compute or storage resources, a network-based service such as >>a >> firewall, a particular piece of content such as a video stream, or some >> combination of all of these. > >This is the general notion of a service, but I think its too broad for >NSC. > >Don't we really only care about services where the actual source & >destination of the traffic are: 1) unaware that the service is being >applied and 2) do not address that service explicitely? I.e., what we >usually think of as middleboxes? Jim> yes. The application of the service function/s should be transparent to the consumer of the service. > >Because end points don't address them, normal packet forwarding >doesn't ensure traffic always goes through the "service". And NSC is >all about being able to describe the existance of such services and >having mechanisms that ensure all traffic (as defined by flows/policy) >flows through the services that are required to be applied. Jim> yes but with the added twist that metadata might be inserted into the traffic flow and be consumed by one or more service functions as it passes through the chain.=20 > >So, NSC is needed to force traffic flows to properly traverse >middlebox "services". But NSC is not needed to have traffic >delivered to say, a storage controller. Right? Jim> correct.=20 > >Thomas > From mohamed.boucadair@orange.com Fri Jun 21 00:00:57 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0AF1121E80AB for ; Fri, 21 Jun 2013 00:00:57 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.178 X-Spam-Level: X-Spam-Status: No, score=-2.178 tagged_above=-999 required=5 tests=[AWL=0.070, BAYES_00=-2.599, HELO_EQ_FR=0.35, UNPARSEABLE_RELAY=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Awv0nfausQS8 for ; Fri, 21 Jun 2013 00:00:53 -0700 (PDT) Received: from relais-inet.francetelecom.com (relais-ias91.francetelecom.com [193.251.215.91]) by ietfa.amsl.com (Postfix) with ESMTP id 9E24521E80B2 for ; Fri, 21 Jun 2013 00:00:52 -0700 (PDT) Received: from omfedm07.si.francetelecom.fr (unknown [xx.xx.xx.3]) by omfedm12.si.francetelecom.fr (ESMTP service) with ESMTP id A326318C8E5; Fri, 21 Jun 2013 09:00:50 +0200 (CEST) Received: from puexch91.nanterre.francetelecom.fr (unknown [10.101.44.48]) by omfedm07.si.francetelecom.fr (ESMTP service) with ESMTP id 83D4B4C071; Fri, 21 Jun 2013 09:00:50 +0200 (CEST) Received: from PUEXCB1B.nanterre.francetelecom.fr ([10.101.44.13]) by puexch91.nanterre.francetelecom.fr ([10.101.44.48]) with mapi; Fri, 21 Jun 2013 09:00:50 +0200 From: To: "adrian@olddog.co.uk" , "nsc@ietf.org" Date: Fri, 21 Jun 2013 09:00:48 +0200 Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: AQHK3KVnq8c9UrscD5I8tUEuvfdA4QOO3MLWmSlTBWCAAQo0YA== Message-ID: <94C682931C08B048B7A8645303FDC9F36EDB2594ED@PUEXCB1B.nanterre.francetelecom.fr> References: <07c001ce6da8$904f9130$b0eeb390$@olddog.co.uk> <94C682931C08B048B7A8645303FDC9F36EDB259223@PUEXCB1B.nanterre.francetelecom.fr> <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> In-Reply-To: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> Accept-Language: fr-FR Content-Language: fr-FR X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: fr-FR Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-PMX-Version: 5.6.1.2065439, Antispam-Engine: 2.7.2.376379, Antispam-Data: 2013.5.21.113319 Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 21 Jun 2013 07:00:57 -0000 Hi Adrian, Please see inline. Cheers, Med >-----Message d'origine----- >De=A0: nsc-bounces@ietf.org [mailto:nsc-bounces@ietf.org] De la part de Ad= rian >Farrel >Envoy=E9=A0: jeudi 20 juin 2013 16:27 >=C0=A0: nsc@ietf.org >Objet=A0: Re: [nsc] What is a service? Why is NSC not SC? > >Thanks to Ken and Med for replies. > >Choosing to respond on Med's thread because he gave... > >> o NLF (Network-Located Function): refers to a function which is >> enabled in the network operated by an administrative entity. NLF >> can be for example: firewall (e.g., [RFC6092]), DPI (Deep Packet >> Inspection), LI (Lawful Intercept) module, NAT44 [RFC3022], NAT64 >> [RFC6146], DS-Lite AFTR [RFC6333], NPTv6 [RFC6296], HOST_ID >> injection, HTTP Header Enrichment function, TCP optimizer, etc. > >Something I pick out of this is that all the example functions are >packet-oriented functions (or possibly flow-oriented) and not application = layer >services. I think this is significant and not accidental :-) [Med] In fact the list includes both (e.g., HTTP Header Enrichment functio= n)...but all these functions are involved in the delivery of connectivity s= ervices.=20 > >Can we capture that idea in a definition? > >I note that it is hinted at in Paul and Jim's BoF proposal text and it is = there >because I asked them why this proposal was aimed at the Routing Area not t= he >Applications Area. To which the answer was "because we want to steer packe= ts or >flows to the place where the next service will be applied." [Med] IMHO, this is a cross-area topic between the Routing area and Interne= t Area. The core concern is still about how to provide differentiated forwa= rding but the technical realization would need some extensions which are mo= re in the perimeter of the Internet Area (e.g., inject a piece of informati= on in packets). > >Adrian > >_______________________________________________ >nsc mailing list >nsc@ietf.org >https://www.ietf.org/mailman/listinfo/nsc From mohamed.boucadair@orange.com Fri Jun 21 00:45:33 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id B4A6721E80B4 for ; Fri, 21 Jun 2013 00:45:33 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.181 X-Spam-Level: X-Spam-Status: No, score=-2.181 tagged_above=-999 required=5 tests=[AWL=0.067, BAYES_00=-2.599, HELO_EQ_FR=0.35, UNPARSEABLE_RELAY=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id nadW0DkIq5wV for ; Fri, 21 Jun 2013 00:45:29 -0700 (PDT) Received: from relais-inet.francetelecom.com (relais-ias91.francetelecom.com [193.251.215.91]) by ietfa.amsl.com (Postfix) with ESMTP id 5E33C11E8168 for ; Fri, 21 Jun 2013 00:45:29 -0700 (PDT) Received: from omfedm07.si.francetelecom.fr (unknown [xx.xx.xx.3]) by omfedm10.si.francetelecom.fr (ESMTP service) with ESMTP id 49FC526440A; Fri, 21 Jun 2013 09:45:28 +0200 (CEST) Received: from PUEXCH61.nanterre.francetelecom.fr (unknown [10.101.44.32]) by omfedm07.si.francetelecom.fr (ESMTP service) with ESMTP id 2EEA44C07C; Fri, 21 Jun 2013 09:45:28 +0200 (CEST) Received: from PUEXCB1B.nanterre.francetelecom.fr ([10.101.44.13]) by PUEXCH61.nanterre.francetelecom.fr ([10.101.44.32]) with mapi; Fri, 21 Jun 2013 09:45:28 +0200 From: To: "Jim Guichard (jguichar)" , "adrian@olddog.co.uk" , "nsc@ietf.org" Date: Fri, 21 Jun 2013 09:45:26 +0200 Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewABC13ID//+4ggP//KpNA Message-ID: <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> References: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> In-Reply-To: <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> Accept-Language: fr-FR Content-Language: fr-FR X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: fr-FR Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-PMX-Version: 5.6.1.2065439, Antispam-Engine: 2.7.2.376379, Antispam-Data: 2013.6.21.61535 Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 21 Jun 2013 07:45:34 -0000 Hi Jim, all, I would even say: these functions are seen as black-boxes and no assumption= is to be made on the outcome of such functions.=20 We tried to capture this and other assumptions in draft-boucadair-* o NLFs are handled as black-boxes; the technical characterization of each NLF is not required. o Not all NLFs can be characterized with a standard definition in terms of technical description, detailed specification, etc. o There is no global nor standard list of NLFs enabled in a given administrative domain. The set of NLFs varies on a per deployment basis. o There is no global nor standard NLF chaining logic. The ordered set of NLFs that need to be activated to deliver a given connectivity service is specific to each administrative entity. o The chaining of NLFs and the criteria to invoke some of them are local to each administrative entity that operates a connectivity network (also called administrative domain). o NLF chaining logic and related policies should not be exposed outside a given administrative domain. o Several NLF chaining logics can be simultaneously enforced within an administrative domain to meet business requirements. Cheers, Med >-----Message d'origine----- >De=A0: nsc-bounces@ietf.org [mailto:nsc-bounces@ietf.org] De la part de Ji= m >Guichard (jguichar) >Envoy=E9=A0: jeudi 20 juin 2013 19:23 >=C0=A0: adrian@olddog.co.uk; nsc@ietf.org >Objet=A0: Re: [nsc] What is a service? Why is NSC not SC? > >Hi Adrian, > >On 6/20/13 10:26 AM, "Adrian Farrel" wrote: > >>Thanks to Ken and Med for replies. >> >>Choosing to respond on Med's thread because he gave... >> >>> o NLF (Network-Located Function): refers to a function which is >>> enabled in the network operated by an administrative entity. NLF >>> can be for example: firewall (e.g., [RFC6092]), DPI (Deep Packet >>> Inspection), LI (Lawful Intercept) module, NAT44 [RFC3022], NAT64 >>> [RFC6146], DS-Lite AFTR [RFC6333], NPTv6 [RFC6296], HOST_ID >>> injection, HTTP Header Enrichment function, TCP optimizer, etc. >> >>Something I pick out of this is that all the example functions are >>packet-oriented functions (or possibly flow-oriented) and not application >>layer >>services. I think this is significant and not accidental :-) > >Jim> actually the intent was that both packet-oriented and application >layer service functions are within scope. It is certainly possible to >envisage service chains that incorporate both of these types of functions >and we should not restrict our work to one or the other but rather let the >use cases dictate. > > >>Can we capture that idea in a definition? >> >>I note that it is hinted at in Paul and Jim's BoF proposal text and it is >>there >>because I asked them why this proposal was aimed at the Routing Area not >>the >>Applications Area. To which the answer was "because we want to steer >>packets or >>flows to the place where the next service will be applied." >> >>Adrian >> >>_______________________________________________ >>nsc mailing list >>nsc@ietf.org >>https://www.ietf.org/mailman/listinfo/nsc > >_______________________________________________ >nsc mailing list >nsc@ietf.org >https://www.ietf.org/mailman/listinfo/nsc From mohamed.boucadair@orange.com Fri Jun 21 01:01:45 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0A53011E8167 for ; Fri, 21 Jun 2013 01:01:45 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.183 X-Spam-Level: X-Spam-Status: No, score=-2.183 tagged_above=-999 required=5 tests=[AWL=0.065, BAYES_00=-2.599, HELO_EQ_FR=0.35, UNPARSEABLE_RELAY=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id wYYoa5NEmy55 for ; Fri, 21 Jun 2013 01:01:41 -0700 (PDT) Received: from relais-inet.francetelecom.com (relais-ias91.francetelecom.com [193.251.215.91]) by ietfa.amsl.com (Postfix) with ESMTP id 6F12321F9EE3 for ; Fri, 21 Jun 2013 01:01:40 -0700 (PDT) Received: from omfedm08.si.francetelecom.fr (unknown [xx.xx.xx.4]) by omfedm10.si.francetelecom.fr (ESMTP service) with ESMTP id AE25A2651EB; Fri, 21 Jun 2013 10:01:39 +0200 (CEST) Received: from PUEXCH11.nanterre.francetelecom.fr (unknown [10.101.44.27]) by omfedm08.si.francetelecom.fr (ESMTP service) with ESMTP id 6AD9723806B; Fri, 21 Jun 2013 10:01:39 +0200 (CEST) Received: from PUEXCB1B.nanterre.francetelecom.fr ([10.101.44.13]) by PUEXCH11.nanterre.francetelecom.fr ([10.101.44.27]) with mapi; Fri, 21 Jun 2013 10:01:39 +0200 From: To: Benson Schliesser , "adrian@olddog.co.uk" Date: Fri, 21 Jun 2013 10:01:37 +0200 Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5t1If55DgDjsYdS26lpw0Q6AX3ewAfuUOA Message-ID: <94C682931C08B048B7A8645303FDC9F36EDB25956A@PUEXCB1B.nanterre.francetelecom.fr> References: <07c001ce6da8$904f9130$b0eeb390$@olddog.co.uk> <94C682931C08B048B7A8645303FDC9F36EDB259223@PUEXCB1B.nanterre.francetelecom.fr> <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <51C32FE2.90108@queuefull.net> In-Reply-To: <51C32FE2.90108@queuefull.net> Accept-Language: fr-FR Content-Language: fr-FR X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: fr-FR Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-PMX-Version: 5.6.1.2065439, Antispam-Engine: 2.7.2.376379, Antispam-Data: 2013.6.21.61535 Cc: "nsc@ietf.org" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 21 Jun 2013 08:01:45 -0000 Dear Benson, Please see inline. Cheers, Med >-----Message d'origine----- >De=A0: nsc-bounces@ietf.org [mailto:nsc-bounces@ietf.org] De la part de Be= nson >Schliesser >Envoy=E9=A0: jeudi 20 juin 2013 18:38 >=C0=A0: adrian@olddog.co.uk >Cc=A0: nsc@ietf.org >Objet=A0: Re: [nsc] What is a service? Why is NSC not SC? > >Hi, Adrian. > >On 2013-06-20 10:26 AM, Adrian Farrel wrote: >> Something I pick out of this is that all the example functions are >> packet-oriented functions (or possibly flow-oriented) and not >> application layer services. I think this is significant and not >> accidental :-) Can we capture that idea in a definition? > >We do need to agree on what a "network service" is, as distinguished >from non-network services. I suspect that some people intuitively lump >together many things as network services, that should really be further >divided into "Endpoint Network Services" and "Inline Network Services". >(Hopefully somebody can come up with better category names; I'm just >trying to make the general point.) > >For example, DNS and DHCP could be examples of Endpoint Network >Services, while Firewall or NAT could be examples of Inline Network >Services. The key difference, as both you and Thomas Narten have pointed >out, is whether packets/flows are terminating (at the Endpoint) or being >forwarded (through the Inline service). [Med] This distinction is IMHO orthogonal to the chaining itself. You may = envisage for both categories of services you listed, a need to invoke multi= ple sets of network-located functions to deliver a given *-Network Service.= For instance, the set of functions to be invoked during network attachment= phase would be distinct than the one to be followed once the network attac= hment is succeeded. In both cases, appropriate forwarding actions will need= to be undertaken under some conditions. The important points IMHO are: * these operations are about achieving differentiated forwarding * actions to achieve this differentiated forwarded are enforced in the netw= ork side * the functions to be invoked (together with the invocation logic) is polic= y-based and local to the network * end user (devices) are not aware of the chaining logic. All what is about is how to construct forwarding maps so the traffic will p= rocessed following policies local to a given network.=20 > >Since we're focused on Service Chaining, this group needs to investigate >how packets/flows are routed through Inline Network Services. It seems >pretty obvious that we already have the means to get packets to the >correct endpoint, and so I assume Endpoint Network Services are out of >scope when we talk about "Network Services" here. (But, of course, >somebody could convince me otherwise.) > >I'm hesitant to categorize based on whether something works at the >"application layer", because I can imagine Inline Network Services that >do so yet still need Service Chaining routing etc. > >Cheers, >-Benson > >_______________________________________________ >nsc mailing list >nsc@ietf.org >https://www.ietf.org/mailman/listinfo/nsc From kgray@juniper.net Fri Jun 21 11:16:15 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id BFDE821F9301 for ; Fri, 21 Jun 2013 11:16:14 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -0.967 X-Spam-Level: X-Spam-Status: No, score=-0.967 tagged_above=-999 required=5 tests=[AWL=-0.501, BAYES_00=-2.599, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_LOW=-1, UNRESOLVED_TEMPLATE=3.132] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 85eAbQXpD+6O for ; Fri, 21 Jun 2013 11:16:07 -0700 (PDT) Received: from co1outboundpool.messaging.microsoft.com (co1ehsobe001.messaging.microsoft.com [216.32.180.184]) by ietfa.amsl.com (Postfix) with ESMTP id 5396D21E812D for ; Fri, 21 Jun 2013 11:16:07 -0700 (PDT) Received: from mail180-co1-R.bigfish.com (10.243.78.231) by CO1EHSOBE013.bigfish.com (10.243.66.76) with Microsoft SMTP Server id 14.1.225.23; Fri, 21 Jun 2013 18:16:06 +0000 Received: from mail180-co1 (localhost [127.0.0.1]) by mail180-co1-R.bigfish.com (Postfix) with ESMTP id 6D38C84042E for ; Fri, 21 Jun 2013 18:16:06 +0000 (UTC) X-Forefront-Antispam-Report: CIP:66.129.224.51; KIP:(null); UIP:(null); IPV:NLI; H:P-EMHUB02-HQ.jnpr.net; RD:none; EFVD:NLI X-SpamScore: 0 X-BigFish: PS0(zzc85fhzz1f42h1ee6h1de0h1fdah1202h1e76h1d1ah1d2ah1fc6hzz18c673h8275bhz2fh2a8h683h839hbe3he5bhf0ah1288h12a5h12bdh137ah1441h1504h1537h153bh162dh1631h1758h18e1h1946h19b5h1ad9h1b0ah1bceh1d0ch1d2eh1d3fh1dc1h1dfeh1dffh1e1dh1155h) Received-SPF: pass (mail180-co1: domain of juniper.net designates 66.129.224.51 as permitted sender) client-ip=66.129.224.51; envelope-from=kgray@juniper.net; helo=P-EMHUB02-HQ.jnpr.net ; -HQ.jnpr.net ; X-Forefront-Antispam-Report-Untrusted: CIP:157.56.237.149; KIP:(null); UIP:(null); (null); H:BY2PRD0511HT004.namprd05.prod.outlook.com; R:internal; EFV:INT Received: from mail180-co1 (localhost.localdomain [127.0.0.1]) by mail180-co1 (MessageSwitch) id 1371838564434446_599; Fri, 21 Jun 2013 18:16:04 +0000 (UTC) Received: from CO1EHSMHS023.bigfish.com (unknown [10.243.78.231]) by mail180-co1.bigfish.com (Postfix) with ESMTP id 5E2B1A20050 for ; Fri, 21 Jun 2013 18:16:04 +0000 (UTC) Received: from P-EMHUB02-HQ.jnpr.net (66.129.224.51) by CO1EHSMHS023.bigfish.com (10.243.66.33) with Microsoft SMTP Server (TLS) id 14.1.225.23; Fri, 21 Jun 2013 18:16:04 +0000 Received: from P-CLDFE02-HQ.jnpr.net (172.24.192.60) by P-EMHUB02-HQ.jnpr.net (172.24.192.36) with Microsoft SMTP Server (TLS) id 8.3.213.0; Fri, 21 Jun 2013 11:16:03 -0700 Received: from o365mail.juniper.net (207.17.137.224) by o365mail.juniper.net (172.24.192.60) with Microsoft SMTP Server id 14.1.355.2; Fri, 21 Jun 2013 11:16:03 -0700 Received: from DB8EHSOBE019.bigfish.com (213.199.154.188) by o365mail.juniper.net (207.17.137.224) with Microsoft SMTP Server (TLS) id 14.1.355.2; Fri, 21 Jun 2013 11:28:00 -0700 Received: from mail73-db8-R.bigfish.com (10.174.8.225) by DB8EHSOBE019.bigfish.com (10.174.4.82) with Microsoft SMTP Server id 14.1.225.23; Fri, 21 Jun 2013 18:16:01 +0000 Received: from mail73-db8 (localhost [127.0.0.1]) by mail73-db8-R.bigfish.com (Postfix) with ESMTP id 0D8FF500132 for ; Fri, 21 Jun 2013 18:16:01 +0000 (UTC) Received: from mail73-db8 (localhost.localdomain [127.0.0.1]) by mail73-db8 (MessageSwitch) id 1371838497385011_28920; Fri, 21 Jun 2013 18:14:57 +0000 (UTC) Received: from DB8EHSMHS020.bigfish.com (unknown [10.174.8.227]) by mail73-db8.bigfish.com (Postfix) with ESMTP id DC995900054 for ; Fri, 21 Jun 2013 18:14:41 +0000 (UTC) Received: from BY2PRD0511HT004.namprd05.prod.outlook.com (157.56.237.149) by DB8EHSMHS020.bigfish.com (10.174.4.30) with Microsoft SMTP Server (TLS) id 14.16.227.3; Fri, 21 Jun 2013 18:14:12 +0000 Received: from BY2PRD0511MB403.namprd05.prod.outlook.com ([169.254.4.32]) by BY2PRD0511HT004.namprd05.prod.outlook.com ([10.255.129.39]) with mapi id 14.16.0324.000; Fri, 21 Jun 2013 18:14:01 +0000 From: Ken Gray To: "nsc@ietf.org" Thread-Topic: draft-quinn-nsc-problem-statement Thread-Index: AQHObqsaqmyiwrYrfkeDlTp72ZjuUA== Date: Fri, 21 Jun 2013 18:14:00 +0000 Message-ID: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.14.0.111121 x-originating-ip: [10.255.129.4] Content-Type: multipart/alternative; boundary="_000_CDEA104415D261kgrayjunipernet_" MIME-Version: 1.0 X-FOPE-CONNECTOR: Id%0$Dn%*$RO%0$TLS%0$FQDN%$TlsDn% X-FOPE-CONNECTOR: Id%12219$Dn%IETF.ORG$RO%2$TLS%5$FQDN%onpremiseedge-1018244.customer.frontbridge.com$TlsDn%o365mail.juniper.net X-OriginatorOrg: juniper.net Subject: [nsc] draft-quinn-nsc-problem-statement X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 21 Jun 2013 18:16:15 -0000 --_000_CDEA104415D261kgrayjunipernet_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Will the GSCP mentioned in Section 3 be described in a forthcoming document= ? --_000_CDEA104415D261kgrayjunipernet_ Content-Type: text/html; charset="us-ascii" Content-ID: Content-Transfer-Encoding: quoted-printable
Will the GSCP mentioned in Section 3 be described in a forthcoming doc= ument?
--_000_CDEA104415D261kgrayjunipernet_-- From rraszuk@gmail.com Sun Jun 23 13:07:57 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 747FA21F9F08 for ; Sun, 23 Jun 2013 13:07:57 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -1.978 X-Spam-Level: X-Spam-Status: No, score=-1.978 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, FM_FORGED_GMAIL=0.622, NO_RELAYS=-0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 5sWupHAWdckY for ; Sun, 23 Jun 2013 13:07:56 -0700 (PDT) Received: from mail-ie0-x22d.google.com (mail-ie0-x22d.google.com [IPv6:2607:f8b0:4001:c03::22d]) by ietfa.amsl.com (Postfix) with ESMTP id C05C721F9F03 for ; Sun, 23 Jun 2013 13:07:56 -0700 (PDT) Received: by mail-ie0-f173.google.com with SMTP id k13so23238673iea.18 for ; Sun, 23 Jun 2013 13:07:56 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:sender:date:x-google-sender-auth:message-id:subject :from:to:content-type; bh=UFju46JXRXcyt6ywxuOJ3HxnqRE+lsDRNAe5EZCKjus=; b=f44/xLg6uMZyuyaDiW2jzGku3e+EL3wuEUIz8hT8Nun3jVXg0Sv9Gd4swENHagClfy C3THWvslgBOMxRcjcjM0Xig2qdbB+NR2/P9HAjnQi5VRCQPaG8O2S5qmsT2+p8uWYkHP VA9BWGzsjmwLgdyPWAVtqVCZfqF7Dls80e48rhQgKkGswUaFjSt2/+tA6vlC3drCu1Sz pM25x8T+DBlv48T5zolvzqLJ6qaXot4jHKI4bxS8UdXCw5D/10Fed9cZQRFwzHMSqdSr ibvJlsJCibMhNaUe+DPdbogLFtlUKgkJGPFxgKg07CheKE3u4ovChTTHnGs0p7eOWqb3 fOUw== MIME-Version: 1.0 X-Received: by 10.42.53.129 with SMTP id n1mr10214758icg.28.1372018075327; Sun, 23 Jun 2013 13:07:55 -0700 (PDT) Sender: rraszuk@gmail.com Received: by 10.64.28.168 with HTTP; Sun, 23 Jun 2013 13:07:55 -0700 (PDT) Date: Sun, 23 Jun 2013 22:07:55 +0200 X-Google-Sender-Auth: Bw-XEWEadGgPgkjq_o412HliiYs Message-ID: From: Robert Raszuk To: "nsc@ietf.org" Content-Type: text/plain; charset=ISO-8859-1 Subject: [nsc] Network Service Header - draft-quinn-nsh-00.txt X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 23 Jun 2013 20:07:57 -0000 Hi, I have a basic question reg Network Service Header draft. Don't we already have sufficient room in v6 header (including it's extension header) to embed services in it ? Do we need to invent new header each time someone proposes a new approach ;) ? Just like we can get service discovery accomplished for free when we embed service in its reachability advertisement: http://telematics.tm.kit.edu/publications/Files/491/ipv6-app-contest-2011-paper.pdf Best regards, R. From rraszuk@gmail.com Sun Jun 23 13:16:13 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id B497321F9F2C for ; Sun, 23 Jun 2013 13:16:13 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -1.078 X-Spam-Level: X-Spam-Status: No, score=-1.078 tagged_above=-999 required=5 tests=[AWL=-0.900, BAYES_00=-2.599, FM_FORGED_GMAIL=0.622, J_CHICKENPOX_13=0.6, J_CHICKENPOX_14=0.6, J_CHICKENPOX_47=0.6, NO_RELAYS=-0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id uBkWC2+T2Y1g for ; Sun, 23 Jun 2013 13:16:13 -0700 (PDT) Received: from mail-ie0-x232.google.com (mail-ie0-x232.google.com [IPv6:2607:f8b0:4001:c03::232]) by ietfa.amsl.com (Postfix) with ESMTP id 3B15021F9F2A for ; Sun, 23 Jun 2013 13:16:13 -0700 (PDT) Received: by mail-ie0-f178.google.com with SMTP id u16so23069553iet.37 for ; Sun, 23 Jun 2013 13:16:12 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:sender:in-reply-to:references:date :x-google-sender-auth:message-id:subject:from:to:cc:content-type; bh=lw/jy7hUcD/N9u6EXOTpqNu/Jl2c/vBzYASEwpxwJ1g=; b=p+GrbPEcSshFNtN7UlJAxxzqE5pI8epwN7gl63OlSp97FZC7Ge3thtXlV7KQY08vKa gfoK+zDhBgBPHBhrcESQ/rkEFwZTOZ0Fs/MiH8qQy0JX+DvVyWxuprd4m9Uil/siWVBj tjSRknrMpFIHuieTAG/0nIdqfG3vbSnSUBV4QtK6TwpeIjVdlixnK+HQK/hu9xkNFxsC /6dxKKFyfWANSK2H+twAnfTop3PhNBaua9opzsfn5TLXcySKFsXRacmW3LZCmQpfER0M p/lwGWTt1T6XfqUBUlSTsOY8+K1qJ4Gs6P/Q+COg+oAqPYmUE4w2h+RsX97ECkfvNMUk cXYA== MIME-Version: 1.0 X-Received: by 10.50.18.5 with SMTP id s5mr1840213igd.6.1372018572778; Sun, 23 Jun 2013 13:16:12 -0700 (PDT) Sender: rraszuk@gmail.com Received: by 10.64.28.168 with HTTP; Sun, 23 Jun 2013 13:16:12 -0700 (PDT) In-Reply-To: <07c001ce6da8$904f9130$b0eeb390$@olddog.co.uk> References: <07c001ce6da8$904f9130$b0eeb390$@olddog.co.uk> Date: Sun, 23 Jun 2013 22:16:12 +0200 X-Google-Sender-Auth: uAZzM1FEg9ps_8TEzn4DWbICt8E Message-ID: From: Robert Raszuk To: adrian@olddog.co.uk Content-Type: text/plain; charset=ISO-8859-1 Cc: nsc@ietf.org Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 23 Jun 2013 20:16:13 -0000 Hi Adrian, Great question ! (Just catching up ...) For me the definition is extremely simple ... Networks = services + transport. So I somehow disagree with the notion of draft-quinn-nsc-problem-statement that services are just data center thing or that those are just firewalls, loadbalancers or worse carrier grade NATs. For me services is anything BGP carries - that includes: L3VPN regular PE, ASBR, L2VPNs, EVPNs, virtual hub PE, Internet, etc ... Those with some history of L3VPNs recall very well the struggle to chain firewalls on customer sites to make their VPN sites divided into DMZ and secured planes. And just to cut some comments short .. this is not boiling the ocean. Very simply service marking can accomplish any service discovery and chaining (ability for specific data patter to be forced to go via such service node+context within that node). Best, R. > On Thu, Jun 20, 2013 at 1:23 PM, Adrian Farrel wrote: > Hi, > > I know this cuts into use cases, but I think it will turn out to be important > for clarity and scope. > > There is a risk that "we all know what we mean", but that no-one clearly > articulates what "network service chaining" is. (It is a little like NFV where > the definition also seems to be hard to pin down.) > > I don't want to rat-hole too much on definitions, but it would help if we can > draw some lines. When is "service chaining" not "network service chaining"? > > Thanks, > Adrian > > _______________________________________________ > nsc mailing list > nsc@ietf.org > https://www.ietf.org/mailman/listinfo/nsc From diego@tid.es Sun Jun 23 15:59:34 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id E4C9621F9D64 for ; Sun, 23 Jun 2013 15:59:34 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.45 X-Spam-Level: X-Spam-Status: No, score=-6.45 tagged_above=-999 required=5 tests=[AWL=0.149, BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id El+ZYXnL+gim for ; Sun, 23 Jun 2013 15:59:30 -0700 (PDT) Received: from tidos.tid.es (tidos.tid.es [195.235.93.44]) by ietfa.amsl.com (Postfix) with ESMTP id 2E96821F9D49 for ; Sun, 23 Jun 2013 15:59:29 -0700 (PDT) Received: from sbrightmailg01.hi.inet (sbrightmailg01.hi.inet [10.95.64.104]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTP id <0MOV002A0BV4CL@tid.hi.inet> for nsc@ietf.org; Mon, 24 Jun 2013 00:59:28 +0200 (MEST) Received: from tid (tid.hi.inet [10.95.64.10]) by sbrightmailg01.hi.inet (Symantec Messaging Gateway) with SMTP id 59.56.03142.0DD77C15; Mon, 24 Jun 2013 00:59:28 +0200 (CEST) Received: from correo.tid.es (mailhost.hi.inet [10.95.64.100]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTP id <0MOV0029UBV4CL@tid.hi.inet> for nsc@ietf.org; Mon, 24 Jun 2013 00:59:28 +0200 (MEST) Received: from EX10-MB2-MAD.hi.inet ([169.254.2.38]) by EX10-HTCAS5-MAD.hi.inet ([::1]) with mapi id 14.02.0328.009; Mon, 24 Jun 2013 00:58:28 +0200 Date: Sun, 23 Jun 2013 22:58:27 +0000 From: "Diego R. Lopez" In-reply-to: <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> X-Originating-IP: [10.95.64.115] To: " " Message-id: Content-id: MIME-version: 1.0 Content-type: text/plain; charset=iso-8859-1 Content-language: en-US Content-transfer-encoding: quoted-printable Accept-Language: en-US, es-ES Thread-topic: [nsc] What is a service? Why is NSC not SC? Thread-index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewAAIKxYAABiYCAAAeHsEAAISBAQA= X-AuditID: 0a5f4068-b7f128e000000c46-67-51c77dd0b577 X-MS-Has-Attach: X-MS-TNEF-Correlator: X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFmpnkeLIzCtJLcpLzFFi42Lhinfg0r1QezzQYMZ/CYvj+/azOTB6LFny kymAMYrLJiU1J7MstUjfLoEr4+i9hSwFh9Qqbt09wtLAeF+ui5GTQ0LARKJvSS87hC0mceHe erYuRi4OIYGNjBI3D99ghHC+MUrcu/ycGcKZxigxd/F5NpAWFgFVifWHToG1swHZj5p/g9nC AhYS227+YwKxOQXiJL4sms8IsUJB4s+5xyxdjBwcIgKpEtvWWYPMZBZoYJTomHaFFaSGV8Bb 4tzGA2D1zAJmEiv/zmSHiAtK/Jh8jwUiriPR+/0bM4QtLtHcehMqri3x5N0FsDmMArIS7+bP Z4XYZSnR8AnsYxEBP4lfU7YxQ5wjILFkz3koW1Ti5eN/rBA/nmeUmHGhmWUCo8QsJGfMQnLG LCRnzEJyxiwkZyxgZF3FKFacVJSZnlGSm5iZk25gqJeRqZeZl1qyiRESeRk7GJfvVDnEKMDB qMTD+yD/eKAQa2JZcWXuIUYJDmYlEd5eK6AQb0piZVVqUX58UWlOavEhRiYOTqkGxkm2mySU 0jzj1q1s/3KyPWbRnmlcN+XXczmGu8csWezFzVNud+vTTo3JC4u9PI7eiWB90vlm5r+sTVnm P7pDfq/gCTGbZHtpcpPirZP5Nwrzzp7tPhMwbZZ3+hWNaI70K7sKVP/q2/qfTl61qCEmMTXx WNFS9rNhen/q3zClHOw+yXFN8fK7hUosxRmJhlrMRcWJAD+iHoKaAgAA References: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> Cc: "nsc@ietf.org" , "Jim Guichard \(jguichar\)" , "adrian@olddog.co.uk" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 23 Jun 2013 22:59:35 -0000 Hi all, Looking at this thread, combining the distinction between "function" and "s= ervice" provided by Benson, and the assumptions Med lists in what relates t= o how policy applies to them within an administrative domain, I am in favor= of talking of functions that are combined (chained) to build services, at = any recursion level: a service can be composed (chained) as a function to b= uild a more complex service. Probably the Solomonic proposal from Jim is the one making more sense... Be goode, On 21 Jun 2013, at 09:45 , wrote: > Hi Jim, all, > > I would even say: these functions are seen as black-boxes and no assumpti= on is to be made on the outcome of such functions. > > We tried to capture this and other assumptions in draft-boucadair-* > > o NLFs are handled as black-boxes; the technical characterization of > each NLF is not required. > o Not all NLFs can be characterized with a standard definition in > terms of technical description, detailed specification, etc. > o There is no global nor standard list of NLFs enabled in a given > administrative domain. The set of NLFs varies on a per deployment > basis. > o There is no global nor standard NLF chaining logic. The ordered > set of NLFs that need to be activated to deliver a given > connectivity service is specific to each administrative entity. > o The chaining of NLFs and the criteria to invoke some of them are > local to each administrative entity that operates a connectivity > network (also called administrative domain). > o NLF chaining logic and related policies should not be exposed > outside a given administrative domain. > o Several NLF chaining logics can be simultaneously enforced within > an administrative domain to meet business requirements. > > Cheers, > Med > >> -----Message d'origine----- >> De : nsc-bounces@ietf.org [mailto:nsc-bounces@ietf.org] De la part de Ji= m >> Guichard (jguichar) >> Envoy=E9 : jeudi 20 juin 2013 19:23 >> =C0 : adrian@olddog.co.uk; nsc@ietf.org >> Objet : Re: [nsc] What is a service? Why is NSC not SC? >> >> Hi Adrian, >> >> On 6/20/13 10:26 AM, "Adrian Farrel" wrote: >> >>> Thanks to Ken and Med for replies. >>> >>> Choosing to respond on Med's thread because he gave... >>> >>>> o NLF (Network-Located Function): refers to a function which is >>>> enabled in the network operated by an administrative entity. NLF >>>> can be for example: firewall (e.g., [RFC6092]), DPI (Deep Packet >>>> Inspection), LI (Lawful Intercept) module, NAT44 [RFC3022], NAT64 >>>> [RFC6146], DS-Lite AFTR [RFC6333], NPTv6 [RFC6296], HOST_ID >>>> injection, HTTP Header Enrichment function, TCP optimizer, etc. >>> >>> Something I pick out of this is that all the example functions are >>> packet-oriented functions (or possibly flow-oriented) and not applicati= on >>> layer >>> services. I think this is significant and not accidental :-) >> >> Jim> actually the intent was that both packet-oriented and application >> layer service functions are within scope. It is certainly possible to >> envisage service chains that incorporate both of these types of function= s >> and we should not restrict our work to one or the other but rather let t= he >> use cases dictate. >> >> >>> Can we capture that idea in a definition? >>> >>> I note that it is hinted at in Paul and Jim's BoF proposal text and it = is >>> there >>> because I asked them why this proposal was aimed at the Routing Area no= t >>> the >>> Applications Area. To which the answer was "because we want to steer >>> packets or >>> flows to the place where the next service will be applied." >>> >>> Adrian >>> >>> _______________________________________________ >>> nsc mailing list >>> nsc@ietf.org >>> https://www.ietf.org/mailman/listinfo/nsc >> >> _______________________________________________ >> nsc mailing list >> nsc@ietf.org >> https://www.ietf.org/mailman/listinfo/nsc > _______________________________________________ > nsc mailing list > nsc@ietf.org > https://www.ietf.org/mailman/listinfo/nsc -- "Esta vez no fallaremos, Doctor Infierno" Dr Diego R. Lopez Telefonica I+D http://people.tid.es/diego.lopez/ e-mail: diego@tid.es Tel: +34 913 129 041 Mobile: +34 682 051 091 ----------------------------------------- ________________________________ Este mensaje se dirige exclusivamente a su destinatario. Puede consultar nu= estra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el enl= ace situado m=E1s abajo. This message is intended exclusively for its addressee. We only send and re= ceive email on the basis of the terms set out at: http://www.tid.es/ES/PAGINAS/disclaimer.aspx From diego@tid.es Mon Jun 24 01:18:42 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7206B21F9CB8 for ; Mon, 24 Jun 2013 01:18:38 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: 0 X-Spam-Level: X-Spam-Status: No, score=x tagged_above=-999 required=5 tests=[] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id sA0wpVRHVrL0 for ; Mon, 24 Jun 2013 01:18:38 -0700 (PDT) Received: from tidos.tid.es (tidos.tid.es [195.235.93.44]) by ietfa.amsl.com (Postfix) with ESMTP id A073C21E8088 for ; Mon, 24 Jun 2013 01:18:19 -0700 (PDT) Received: from sbrightmailg01.hi.inet (sbrightmailg01.hi.inet [10.95.64.104]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTP id <0MOW009I41QEEQ@tid.hi.inet> for nsc@ietf.org; Mon, 24 Jun 2013 10:18:18 +0200 (MEST) Received: from tid (tid.hi.inet [10.95.64.10]) by sbrightmailg01.hi.inet (Symantec Messaging Gateway) with SMTP id E6.7C.03142.9C008C15; Mon, 24 Jun 2013 10:18:17 +0200 (CEST) Received: from correo.tid.es (mailhost.hi.inet [10.95.64.100]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTP id <0MOW009KJ1QHEQ@tid.hi.inet> for nsc@ietf.org; Mon, 24 Jun 2013 10:18:17 +0200 (MEST) Received: from EX10-MB2-MAD.hi.inet ([169.254.2.38]) by EX10-HTCAS6-MAD.hi.inet ([::1]) with mapi id 14.02.0328.009; Mon, 24 Jun 2013 10:18:16 +0200 Date: Mon, 24 Jun 2013 08:17:15 +0000 From: "Diego R. Lopez" In-reply-to: <08db01ce6df7$38e56330$aab02990$@olddog.co.uk> X-Originating-IP: [10.95.64.115] To: "nsc@ietf.org" , "adrian@olddog.co.uk" Message-id: MIME-version: 1.0 Content-type: multipart/mixed; boundary="Boundary_(ID_r/Imt8r/JUrLo5oMdJhX8Q)" Content-language: en-US Accept-Language: en-US, es-ES Thread-topic: [nsc] Whither NSC? Thread-index: Ac5t9y4GPlWSXJZKTc2hRZi75qXMZQCq27kA X-AuditID: 0a5f4068-b7f128e000000c46-7f-51c800c95e95 X-MS-Has-Attach: yes X-MS-TNEF-Correlator: X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFnrEIsWRmVeSWpSXmKPExsXCFe/ApXuS4USgQeNxfYvj+/azOTB6LFny kymAMYrLJiU1J7MstUjfLoErY/W2CawFn9YLVOxZkdXA2POJv4uRg0NCwETi9wOxLkZOIFNM 4sK99WxdjFwcQgIbGSWuLWtiAkkICXxjlDiykQ8iMY1R4v2ju+wgCRYBVYnVbfPBitiA7EfN v9lBhgoLyEvc31cNEuYUsJZYdGYnI8QCBYk/5x6zgJSICPhJbO+wBQkzC0RLNO3ezApi8wp4 S0ydv5oRwhaU+DH5HgtETbrEpfN72CBscYnm1ptgcUYBWYl38+ezQoxUkJjXKQlhGknsaEuB WCoi8fDiaTYIW1Ti5eN/rBBPWUncWP2NaQKj2Cwky2YhWTYLybJZQFOZBfIl9v2sgAjrSCzY /QmqRFti2cLXzDD2mQOPmSBsM4lrhxuYMNVUSJy5e5cdwtaV2Hp8KussYNAyC2xnlDh4agcr TPPlfeugGhQlpnQ/ZF/AyLeKUaw4qSgzPaMkNzEzJ93AUC8jUy8zL7VkEyMkLWTsYFy+U+UQ owAHoxIP74P844FCrIllxZW5hxhVgAY92rD6AqMUS15+XqqSCO+hK0Bp3pTEyqrUovz4otKc 1OJDjEwcnFINjDLmMrOat0lfaLw965V28Mkz4RedT6pZ5syuC3kaL3ar6JH3Aa/OD2H1K/Jm 33ns+lG6e6XI9RMd3V8vGWzcyFhZKats8TmK2V38o/UC9mnbkq8cXXyHVZBbb8ZP1yUZc41Y rh6Ze6PAe9rEUNMf8SdtA+ROySzbzd9/Lje8IXl1U/WiO3a3viixFGckGmoxFxUnAgA6o3eZ 9QIAAA== References: <08db01ce6df7$38e56330$aab02990$@olddog.co.uk> Cc: David E Mcdysan , "stbryant@cisco.com" Subject: Re: [nsc] Whither NSC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 24 Jun 2013 08:18:47 -0000 --Boundary_(ID_r/Imt8r/JUrLo5oMdJhX8Q) Content-type: multipart/alternative; boundary="Boundary_(ID_3TTyu78ampbxzCwmhtoeUA)" --Boundary_(ID_3TTyu78ampbxzCwmhtoeUA) Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7BIT Hi, Here you are a document on service chaining being discussed in the ETSI NFV group. I am cc-ing Dave, as he is its main perpetrator and I am not sure whether he is in the list. For the BoF, I'd be more than happy to make an introduction to NFV and the ideas we have discussed so far on this, as well as presenting the use case we are considering right now, related to the composition of personalized services, especially on issues related to security. Be goode, On 20 Jun 2013, at 22:46 , Adrian Farrel wrote: > Hi, > > The IAB and IESG met today to talk about the 15 (yes, one five) BoF proposals on > the table. Of course, NSC was in the pile as well. > > We discussed NSC quite a bit, looking at the proposal text and the material in > the drafts, and also considering the views of some of the I* who are actively > involved in the ETSI NFV group at the moment. > > The was quite a suggestion that this is all too early for a BoF, not least > because ETSI is not scheduled to come out with its results until end of July at > the earliest. It was also suggested that discussing solutions work was very > premature. > > In the end we compromised. We will hold a short (1 hour) non-WG forming BoF to > discuss the service provider requirements and use cases. It is my intention > (and, indeed the I* was pretty forceful on this point) that presentations should > come only from service providers. The vendors' job is to listen and ask > questions! > > I will be appointing chairs soon and then we can start to put the agenda > together. > > Of course, the mailing list remains open and I hope the discussions continue. > > Thanks, > Adrian > > _______________________________________________ > nsc mailing list > nsc@ietf.org > https://www.ietf.org/mailman/listinfo/nsc -- "Esta vez no fallaremos, Doctor Infierno" Dr Diego R. Lopez Telefonica I+D http://people.tid.es/diego.lopez/ e-mail: diego@tid.es Tel: +34 913 129 041 Mobile: +34 682 051 091 ----------------------------------------- ________________________________ Este mensaje se dirige exclusivamente a su destinatario. Puede consultar nuestra pol?tica de env?o y recepci?n de correo electr?nico en el enlace situado m?s abajo. This message is intended exclusively for its addressee. We only send and receive email on the basis of the terms set out at: http://www.tid.es/ES/PAGINAS/disclaimer.aspx --Boundary_(ID_3TTyu78ampbxzCwmhtoeUA) Content-id: <4DAA536FC10B124681603FFFA8A0F3F2@hi.inet> Content-type: text/html; charset=us-ascii Content-transfer-encoding: 7BIT
Hi,

Here you are a document on service chaining being discussed in the ETSI NFV group. I am cc-ing Dave, as he is its main perpetrator and I am not sure whether he is in the list.

For the BoF, I'd be more than happy to make an introduction to NFV and the ideas we have discussed so far on this, as well as presenting the use case we are considering right now, related to the composition of personalized services, especially on issues related to security.

Be goode,



On 20 Jun 2013, at 22:46 , Adrian Farrel wrote:

> Hi,
>
> The IAB and IESG met today to talk about the 15 (yes, one five) BoF proposals on
> the table. Of course, NSC was in the pile as well.
>
> We discussed NSC quite a bit, looking at the proposal text and the material in
> the drafts, and also considering the views of some of the I* who are actively
> involved in the ETSI NFV group at the moment.
>
> The was quite a suggestion that this is all too early for a BoF, not least
> because ETSI is not scheduled to come out with its results until end of July at
> the earliest. It was also suggested that discussing solutions work was very
> premature.
>
> In the end we compromised. We will hold a short (1 hour) non-WG forming BoF to
> discuss the service provider requirements and use cases. It is my intention
> (and, indeed the I* was pretty forceful on this point) that presentations should
> come only from service providers. The vendors' job is to listen and ask
> questions! 
>
> I will be appointing chairs soon and then we can start to put the agenda
> together.
>
> Of course, the mailing list remains open and I hope the discussions continue.
>
> Thanks,
> Adrian
>
> _______________________________________________
> nsc mailing list
> nsc@ietf.org
> https://www.ietf.org/mailman/listinfo/nsc


--
"Esta vez no fallaremos, Doctor Infierno"

Dr Diego R. Lopez
Telefonica I+D
http://people.tid.es/diego.lopez/

e-mail: diego@tid.es
Tel:    +34 913 129 041
Mobile: +34 682 051 091
-----------------------------------------



Este mensaje se dirige exclusivamente a su destinatario. Puede consultar nuestra política de envío y recepción de correo electrónico en el enlace situado más abajo.
This message is intended exclusively for its addressee. We only send and receive email on the basis of the terms set out at:
http://www.tid.es/ES/PAGINAS/disclaimer.aspx
 --Boundary_(ID_3TTyu78ampbxzCwmhtoeUA)-- --Boundary_(ID_r/Imt8r/JUrLo5oMdJhX8Q) Content-id: Content-type: application/vnd.openxmlformats-officedocument.wordprocessingml.document; name="NFV(13)USE004r2 cln.docx" Content-transfer-encoding: base64 Content-disposition: attachment; filename="NFV(13)USE004r2 cln.docx"; size=190794; creation-date="Mon, 24 Jun 2013 08:17:15 GMT"; modification-date="Mon, 24 Jun 2013 08:17:15 GMT" Content-description: NFV(13)USE004r2 cln.docx UEsDBBQABgAIAAAAIQAZsGhy5gEAAAUJAAATAAgCW0NvbnRlbnRfVHlwZXNdLnhtbCCiBAIooAAC AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAC0 Vktv2zAMvg/YfzB0HWylOwzDEKeHrjtuBZZhuyoSHWuzHpCYtvn3o+zUSFMj9ubmYsCWvwdJUdTy +tE02T2EqJ0t2VWxYBlY6ZS225L9WH/JP7IsorBKNM5CyfYQ2fXq7Zvleu8hZoS2sWQ1ov/EeZQ1 GBEL58HSSuWCEUivYcu9kH/EFvj7xeIDl84iWMwxcbDV8jNUYtdgdvtInzsnBGfZTfdfkiqZ8L7R UiAZ5WmVD+J+e9ieALVJwu3CMMbbYUj6PowI0MQTlWN791ad5CA/xF8Qsg0h1trHd4c4vlEBglaQ 3YmAX4WhaPmDC4orJ3eGMlWcT8eAnqsqLaHHJzYfnIQYqbKmKfoVI7Q950PuIjrzyzRcI5i74Hy8 mm2nJ018EFDDeC7szmwgkPvZ6i+S0VOfS0RbkIj7BuLrO+h4J8r/1FjfVhVI6q3xjWFinqwXncQR dlwNECnfU0Sed3w+tvvigXnUwgNsvl/MxRH5qJHKObQOL1H7nnrUBFh1IQ9PzKMWahAKwvwT4EUP dsSj+ilZF9HviCfoW1yLTQMTOu8fm6KiaddSj5pAGrPA2+f8SrQ05yRpVLTHPo3t8B9hP02+hM5p Bk0473tFGqqz8wzpUqFADWjz9hKz+gsAAP//AwBQSwMEFAAGAAgAAAAhAMJgmvP0AAAATgIAAAsA CAJfcmVscy8ucmVscyCiBAIooAACAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAACMksFKAzEQhu+C7xDm3p1tBRFpthcRehOpDzAk092lu0lIRm3f 3kFUXFhrj0lmvvnmJ+vNcRzMG+fSx2BhWdVgOLjo+9BaeNk9Lu7AFKHgaYiBLZy4wKa5vlo/80Ci TaXrUzFKCcVCJ5LuEYvreKRSxcRBX/YxjyR6zC0mcgdqGVd1fYv5NwOaCdNsvYW89Uswu1PSyf+z 437fO36I7nXkIDMjcFqhZMoti4X3mD36r8ZKlQHnbVaX2/y9KY4s5EkIXcy8SFlzytJrsj9C6vKk 1+Wz4pzQzeVC0+Xn4uGjcPDszytRSt9GOPkFzQcAAAD//wMAUEsDBBQABgAIAAAAIQDCQ6xz7wEA ALsIAAAcAAgBd29yZC9fcmVscy9kb2N1bWVudC54bWwucmVscyCiBAEooAABAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAKyWwW7bMAyG7wP6Dobutey0a7ahTjG0zZDDhqHx1t4CRaZjoZZkSEqX vP0YJ3GdtdYuuhggBf/8SNGkr282so5ewFihVUbSOCERKK4LoVYZ+ZVPzz+RyDqmClZrBRnZgiU3 k7MP1w9QM4cv2Uo0NkIVZTNSOdd8odTyCiSzsW5A4UmpjWQOTbOiDePPbAV0lCRX1PQ1yOREM5oV GTGzIk1JlG8bDP1/cV2WgsOd5msJyr0Tg1aoZGqhnlGUmRW4TrbQfKk3MTgrWtDZ/Bv9Mf1N8/kt HSfndw9fp/nOSJKU/nycLfAMrUV6kaSLlyS5jFHgKPpdF8h7v3FgFKsJHchsFDQzYAWYIwGm1dpp jPUfBLgICVBqjfm+AuxtP8BlWADlcrasoc9wcHnL8DEkhcPW7xG0Jm2f/lKEZOBr67R8wpvvujyO aeelwoH00gTtTLWWSzA4UF5pOpfvXoJ2p3XbGkdXV4+97Qs/1JtScKOtLl3MtaT7kbMbNePTcUb3 ER6Fq+7LErh7E7x35OMI2p0WnMOb6LMcPD6Eq5DN+QeW8zcUPacPZBwSZDegFM6sXjE6lw8CN2K4 hQSq+Ifh6PEhfA6JICQuZBQ87EMJhWC0daZxg1/t0PrCX4VwdRiGGB0h6Mkvx+QvAAAA//8DAFBL AwQUAAYACAAAACEAJHih0kxPAABitgEAEQAAAHdvcmQvZG9jdW1lbnQueG1s7H37ctvImt//qco7 oJSqZLxlW7iD1K65AW8+2vV4HNvHJ1XZrVMQCUk4JgkuCFr2SeUJ8mfeI2+Qt0neI7+vuwF049IE KcnyzJmpGksigb58/d1v/Q//+HW9Mr7E2S5JN6/OrJfmmRFvFuky2dy8Ovvjx/mLwZmxy6PNMlql m/jV2bd4d/aPo3//7/7h7mKZLvbreJMbGGKzu7jbLl6d3eb59uL8fLe4jdfR7uU6WWTpLr3OXy7S 9Xl6fZ0s4vO7NFue26Zlst+2WbqIdzvMN4k2X6LdmRhunfYbbR0tioFt0xycr6NkU47RXFG6jTdY 73WaraN89zLNbvBG9nm/fYEVbqM8uUpWSf4N6zP9cpgvr8722eZC7OpFuSt65wILuPiyXhUPY9nd z3IIXPAfxRtZY6Mti+SvTAXI2fLOs3iFBaeb3W2yreB26miAx22xJO2Gpc3ebS23MV8Jnj6HPs2i O5x9MfHdtjFcCzCW/KX1isOBEKpCo/qIltnjRGiIcg19lqDOWaxERr6700BTYdLdFiR4H4J6naX7 bbmrbXK/0S43n8uxiBMcsTLTZ6Qub2131AANXvHhNtrGZ8Z6cXF5s0mz6GqFFQHiBmHk2Qjc6Spd fqOf+dVK/HiXiV/+ZNxd3L06G/q2d4Zf829bvLz8Gp2diwfGwAawRPZXusUjX6LVqzPiUStMenex ++urM5f9so0WeBdM8+5ika5SkLLJ/uNDreLr/PS3r9I8T9env58lN7f3mD7Z7JJl/IfT5+cDfDp1 gHN2dvJRXK3eRN/SPe2Jn9l18jVelqc2iVernyN2yALwOORB2xEXkGn/nk8sjQYUepOmn4uNmG7I Dvw6yXb5+xSDWHT+q0j8VX05SVf7NYnV4vviA/bIJv3DGIJVoM8m/cT/smhDfA0lxr7OkiVh4w1+ YgzMhllty7P55pWPLdt3Wj52rGHLp95gKM1XTJNnmAEqwfI94bPn2dNJQFtgH71jSG7PPCcg2Fru xTbKokvsw3XnjudhdvZpHn/N6dNA/FcM8JG9b3n+YDJjlJpnYp/ZH2KBsrfv90TQ8ddokdN7jP4s i+2LQFO8sWAkuhDvL/SETVD6sI02xXguh0e+kHHsoch9k6z48CoVlx8XKMh2Jj6lnVWLYX/xnYEF SWAOpvO56wYTPZi7j84ZB1YItU6c5zS+jvarnE56MHNMjtt00u/oI/EwwfkKNEC60oc8yogGkxJ1 ozUO688f04XjWYHtObZf7L35BqOFjfSGMzCHtj/gb2TFqjiWSbPnow+MKhhtgMgBni2nEyABwwmG C+UA3ZgbOv4s5MArMNd2g8HUHRKSA9D9Mbcd/xhhYiN1wVKdLb6rEK3ECVlclB8+PP64vjUfDs1H wB+VU9TxpzyalrP9mOSr+O8uuk+YSKOV2AM3IIZTB3ad2AVLlOjrac/Ac+aO7RIVSgjn2JbtBAOb 9qNnv900fPIZkN11seMKzTaLd3H2JT4bXW62ELh5arydfzKU48ESlWWGTugz4pakhES/gmVn83ST 7/DibbIBz4khNMNdwlQvEHHB19vX8sddbEyiXbzTLiSYmoEzLkHYgmwvjA/YHExRY3ILa1G3q5nt zAOfi6n2RRk/fXo71y4onFuDiXaMeZrdRRnZ3NqB5MWMXmfR9nb3TFn8yTxxalu+SRqtJGbcycwy x/4BNtFHmnPp7ZrE4hmnLuR1K0H/GNwTOke4Sm5KZWEBR0ecFTsQGsdWBVg4MANvNuVg7BIi3XJ5 aJlOUCGuJJe7aVpAVLDGVnj+0AzyBDA7U2cwB5c8FczToet4J4P5vpqGPx6OfWdGIkti/Id15Hbh 96ullcAcmNbUE2rY8bRyqvyDmyi9nmVZqTDcZNGaKbSH9c/rLF0r7BYaaNtws42wSWmWdmHIfKZN YWt8SPfZ4jeoB51A5rBy7Nl0ejKZd2OIqiAco6V+irPkr6mqMJwsc+02O8Q05wPfITXqN8MdiGO2 y0sHSrDnky9Es9luedl9wt3ykiSkhixHE2in8Dn87RgigeVPbNOtaXlWELpjb858Z6UhYg89d4yg iGBqkn6ip6hWJtmT5y6jL/HL9WL5bRdt/jPCVER+FFY4gg9zAgWvVpQ1058P7fG4tu+6HGbIV9u3 YlC0AEN9XGYvlbJ2Xy3CmsLYGg5Ii9GQTlM3/7G1iHzRcN8yR2nNj1K4zrq9t3ycTsZjzobjqT+p Ker2zJoHLoPpAeP7b1FR7wbmeDCBNi6Mx+M1ufsB876EZNvOeDiYHZBBvxNSuwT33dl8Yh2C3sNL 8PtIlIQcWkeID62+0OESQnaBMUnX6yTP49+gNt/JC7zJbDoMfyjPcvsJfVdn5qj0W9bQTqOIti/b +FOafTYu83it+glxIKfFQuYDRPFsrkUUsRDTmfvTmV+TjnXN6NfFEjsx1nd8dzA1h3pFqpuD3U96 kWX06/PZdQLTdgaOP5nNTwXm03rm7PlsPhxMetnes7ETzjuj19zfjTDOj+nvPjaPhZ13GSPvPH3H HM/DgAWWNTZJNyk9ljmfJVd7ypIzEGU5KcLIEipKh2GVJ9QRzj0WvMgc+fWmCfXEDRcu+5CHB38Y 3JjGi4SyXhWhjP1UkrSTPVMuze8IQSBoJL31RIhgOBhY/vCAkvHwzKIth4X0nwuekQLFvUALHg8o nodPHw9RxotIucrwZ4dvf/RfT0AploiFMWu+jqPyFY7nPCcnl5zi0HengyCciZyx4x0Fp2gHpFxt ebRZpBbAh5grgd3dfotUz0WWbHMhr9mTxAjoRYUhEHLjM/q3OP6j0ozc8cS2Ta+Xt0FWMdqVxO8e /DsWv3ryAsubOPMpkgG1zsxuXvBIOvjvYr87tbnvyYazyWA8PBDs7j7Zx1AJp8lusUe9ye+Cv5Gu XjjWmf3ySMQezMeOa4UHIhfdKOFZ/oDndDPpKwWhuoN+VdylU6P77YtfK4REMecUvztJBf8NiF9r GFhgR8Lf1aWBcM/W3474DSxvFgynB6DSTZG/i99TSoO+B6/1zMAFt+wT41aZ5/tDMW71cTnGfSjB 4nIDGwulhr/L32a52PfACWsyHQ8982Qf7Snyl3CiMKTl4hHSCyvDW8KMLtu7qP/p4aL57Qt0eza3 rens5MD7b0CgB0EYTm33gOhqhqqKMoNDBWfBbz1l3XFN1575J6fhPpLs/8FT1o/PU2qmMTELWqQ7 HV1pgEQlL7Cdk+3qU0i/MqLu6wXz7OEwHKLuVGuJNKn2B/GCtRZrnuAP9eee5XjzR0hwPl07+7C/ WifMOWIso/w3mLpCDoBjkwwfmnhRe+nNzTrxWqE5syckyg5kH3Y7xSzPm6JYTYzQ0MqZ43pkKcGJ dsWMdAqhmP35dTqOFp+5RpbJi5sO/RnqaPmoQceoVdxEVFtz9/noxT/tN3HtHWV0FYdVg4RP+QIt ZBxlCDCmSjM8hku1lb/hNIaoNKof04MVs/+6dAsmrtqK0f2JifpG72Qu9reoQXQCMwinw/FkfDIw n1au25OBbVuDXtEtuf3DjyzXu4/KC1G8Yp4czezm4irna3DxggUrLhrOEX+O45xqiP9jtN7+vRGu VumCOVr+dqpZ7Ol8PPbNmmqJYxrPptTP63TZqj+VNu8GK57haQXiYN7H13GG7mpxTZq2HOXbuSLZ KjFdCVSvNgpFoiew5QJLyORPxsHEU3pHluMdaafKWkASp0nZtoI313ZDZzw4YID+umyBTp5hBe54 Pjk9XX7gOEOWMsM1xIeKff3g1vbR1rE3dMxpeLpj7BFYcx+6qmhc9pEypsW5yPt4FX+JNjklgVMO +K7LTcq675xMphBsZaZekROOXhCoWzeP7o/T19H2fTs7lb115MTL8kM1J+rYMDgkjNx2rdVZ0Mkg HBd4GxxqoNUdiiubcp3EIGhZMM/oB++Zt2UOmgIHAhQmI1u+lrLnjd3Qm7kUT+i0W3sxKlJhSguw KRHbKSgcf/j4Ppx8vFDrIdpJibUw46QUXu3yjHqLwatVPVzJ1qAmW/uu5+NtsjMWqCEuk45RqrRN D7WsERyHgYAbWasIehycStR3MN68+OMHWhBOpiCoUVsZUwk+Euvow+UFZEnwY2nsQIx0n348daAr 8x+/p/YjploqRf+Q0IT22ZznPlvKb+MjZzt0Uu27OlM6EKHFbrxgOerptXFwCeixNmX5x+35pw9w sIfKox4Ut0Z7tHVasLZORUfhnufd1UVjT0ybwRFScoseubEBuNYG7V1DWEOyfATo/NRnMLCTGhtp DGU9++OHmWlamfVSGRGUDh0b/ygc2B7PTORME689lMIjqKKgfokDq9/I5iXjheqM7twy596kj0NC Hfe9wm/uuwzLsgOUl9SW4XjmZOgN+oqeoRvMzHHBEOWNM75L/+SjKjvwnNyeN+igu1m22c/1s3HQ ztIMhjT+obNRF/JeOhv1G3mJbWczC9FDs64smD5sDHhXin0yZaF2NtKM1N2TDcGkhDxjBZR2DsZE XUGwRvwVjTWXOwNlNof4lz+3gmAqbNS3v0yMkv4VEgCbZ0fSPrtBzXiOJml+xsfTr2YhlvPsZ9M2 nSG20cZ1aDKJytUlt5O5h+II1+7V+UbAsoW+1G/kg21DJXsyQV5ITb335g6SNR2VvoS3oGVG9ZtD M3rzCWroao4aezpDcqjFPCdl2xF1XJlc1G/kGQ8jb2xktpHFX1gJ0M6Isti4ggxaGvutwGGKUBhW YFyjxfML+sdYC/8a+o9Q3/vdS+PnFO9RIqFxF2GQ5RIDZPGN6OV3C0sgzl7AaotXBuQNmlSvjWXM iw2QALV7boBojHlys8cwDvuDqGcZ51GywlBokRhnbGr5LeMOXiRMA8m2wENX34zI2O3XaM36jeQc jVBMtssh/Nha2eAI7e1jLFshtHYkRC/YwWTA+MghflbjLg/K+VEwMvMamFkvRmZMLjBtx64UXonJ iW8IKUQxSIoykOtVevduv1nkjCCifZ6iOfAins5qf7/lfy//skd/Z7LtLjdL9hH1yiSSv7uIrnFO KA0yWahwlVB7dJtH4ugP3syYZuD6ACEMq0FhtycILb/6uq7AFT0yX519TNZofvk2vjPep+sI1xyg AAuN1Osf89X9Fd8y+8ER3XiFploN1zAsCBVEtQwB6yrDELwgaRvdMK8lnuAmYpt6EgxcCOSaL881 J16IfrOKQDJDtJusyJyjjNpFSCZ09XFG6OIj6Uy3H/Jvq7jY9B/iiPppsubWWLPYVdF8tIDBR1xT gbNaJ+ghP5NakVI377oBhpGL116d/SV68U/vZJtMAh0elDTzECqKxUuFWYfweIPW8vHyHQA6zuLo MzusvOwRYPwHhTgxVDEcKY+I71q2zWTnUVvCGM1+q+UmNdsinahzRfzYpA12KOSd1nllcLNEqzZL vDhAGQoCURiTP/pgEwbwY4+X4CBMNt35CFb4nVfWoSLpFiq1kj0KkRjwTgch67e70y0MJeRz+Lh/ DAjWzTyFFp8IhOg2rIOf1Gz4Ox5sBwai4l70NdYt+WkA2bFk3lZZt9ynwdBRW6fnuhGK/KEpEp9r LtpWbQnycxYyj2i9jmo4912WnFeZhOLhQ8K2KI0+mic3NBOdeD2aNdxL+P2c5smXrnKFOvx9O5y6 0/rNDE4YjCderau7M0H+X10DqmnSkgo7txH6K436uqlTKG0c8MSeSU+jvyS1RIJbFwOXXoTAa6eQ t2keNwS6whh1W5MWoXHTotK+z1J+eofW9JFsURl3t1FuXMWb+DpBl/lFul/BQIqNaHGbwApbqpZP Tbkah+ZwOin1VMWOkdbda2lwWgsbbGfs0tWhqR8BZCPYhTmFIwr3Sh8G4s6QzuW7fWoKBNcu9DIJ V9VvZFxl2sNW8WnCdxLOwkHNEeBP5zNrPKsOg5Tfse1PWXYRV697zcjOjf7JR2c1pt5wAO+2uEBI 7kvK3/s5WS5XnbmKDSzJR1fp115TtfuHpw66b5lMAeqgQAQL2O0Fu7NOdG5ZFXk44LiAd4NcF+Qo iGqrBLsgOG1wAeDyBf43NnGOa84+F/MBnaJ657yS8JsztrMPY5kamzRnWEmeCoThlgmFPqJVYzds Pe3DLFYJtvGCIBFnzxtvFljZXNUInhQkSTGOLjw+iDAVty9QrHOZ3KwJPJfvmM9khlVmgITqoK+x jvnc9l2Rd9O+YOMXtJmI8jRraL+aXRoJwhfMfwM2Vz+tYo9EGjXTuWXbm/36Ks5oX8fTwXEzrRm9 3I8IBKnrQFMSwb3g0nFYP8Uvb14+N3DDV3wXrVbPjbfhx+cUpNtnuBzyubFKI8iVCHrLoj69Zs2E qdN3l88N5/W7d8Y1+Z64F7A+hgxwf+7T7UiaYb9EWZLudwYVVjLCvoHT7y76BvdifWDNKHG+ePmM nIlLOBbTb8w1BWzZ7jOKX7+42ier3NjeftslC5BqtN2uEuy+4U2sEQbugnIcT7f6xhrlzcvMcPRR YhWCNZH7bR3ntylCD8TgIOuWdFkNQvDX3Kt6hUu/buLdOQI58NHhKfA1UDRCjs29VOTG2AOD5f2o RVk/J4z6kbRst/iohYzvR1bycjowvyQruJQXt4zlG/Bzfo4hM8hjik8Ayhv6Bs7sZEO+5/jf9pTV yB+GxnGdsU/y1TfEdVfkyI6MMzGwsaArd17WRBepqYSbqm4QBK7jWOGBFHzm/O3WDdRvDmkjA9Tk zICwahRvDlU9dMis4qoHsVxVy5G9lQLK4mF5RtqikLIfSfbF2ZoBRocTMisc0aVFZwZALEQohQjY zUz4gQMxoit4tuujaYgeYQtwM/TH429DCKc3xEooAYBWp79pCTl2cwQP2fij/1afVgaW9OT1agkv EOkN4reP7PrNq/gG9/YKrZ/BqeN93CqZZx8p5NJ6X5XxfjY3Lj+8/vNHLB9uXWzom/Evt4zLlK9y +0g3S8cqdzFpCLnkCgfOZrUrRRp6ZMdO8hHW2QNqxeucHQzNYO7NBdDlXdZGal9Wu855+HigDvY5 nI5D+dcGk+c4A5ZdW3WpTxKJScvqGBixEK6izoUwNX56O39GklCQA0/zKmVtRHc7NxZTAFgyMvj6 AMMvuANVxNgKFnZGQqS+bvZCB0p+SupPd86Y5XtI1+au4PVj26oPpJsWnBk3b0O1jWuCrT6Ishq0 55nqzQ8yHQwg+SLBdd4lY68PyiGo1ZrH0lwN80zBBJkJdgDZCMulcBlzJiLyBS8rmAG4J7HKBowl MwCXAm2F7FvCk7BB0I1eKQVdel3frQJC1U5tSvGOHfQYU3Pcb+e0SHg/CmkNywq20S5+afyygSSm DehmkE2Y0aeEIWLyV+A+DXyAUHscDwTUboGrvEGbulWYgYkWPTp9MTt4E6G8k3ZQ/3R3m0C5WUff ID8Xq/1SDxpr4uBSU92i3li9Ydu+ojLSj0NMK9UUsH+mHfkwvLjCCxXe2KYZszeY7a8d9fB5HqYe YjsZFbwg2El7ovwI4+z+dyVKZKpDJHkLI0bP0Ju4knSdojKKLp1XmVO7DuoG47E5YXkFh/IgujVC 9RtZI2zxiMFdbzrTac0Hh+I61w1ZL59KBxW2YcF8JI+Y2L74hs0oGZKZho/ghF7wTBgJ0gyAemXQ Cy1cX6ojEuYHZXyoPIHKSQo0PT+Q6SttoEMdEPJai9uH11m6j3X4NZ/a86noAD4CbpNrDSiPLrEN Yc8lYX4L/G+xOTmsD9KTbi0T33GRecjnafiXCuwg7UZZdR/0D0IvtKcse+MQ+qtmzwGDSJwlQ39R bYAfPH6BX5RUjo8REqfoanKRPXO1+hP2hNt/pML8KnsGb/+Il7VbviOuuKabics73J0BrwOvXe3u Om5RxIP90N7Z4So60Wzmev7sTPGm+0iJRcOvAwFAXk4nE0J7CS5bM0Ct9CQGm2y/DMaykao1Zp2x JVTxJ27o4o4xrLPiW+XSm578YlPsm3a+NQpzXl+hGg+cf7PlMWC174kBvPeevCl6lTj1gKqL3K4Q jonT9yRpvjpe/K5wt4WFu03L2qRR89G89GrrWIeULtDBUxszcjbD4uTKyP3gzzC7N/wt5J9Pp0HN H2PO4X7xrSp6TaztfjjVrpXdX1upTuE0OBI6cww5SPvWfD6YuGAnqu+qtWHFQ9O+PZkN5hO35qnz J7hFEL2KetJJgDw5X41xSyKfQ2E0u4bRidjP4tsJyHcc8ZuWbXvT+v2Uto8bjkx2yUkfhiYgLbhf B0Obxkt41XNoD1rD0ndxabzouNKOsIUXhJT+jXCNKHAq/Y+LiLJX4QzR6guCqDjJM2sQSqGxRdoi C4EoQz8K/c9wxZjPqp0kmRLMBr43ttXEdAGcQsxIunCvIygdSDLoKiCxvVNgYQ97ZpPzABA9uoMv MyYFBTkGWbqD81ucgRau8kmO2vQwIcaOIH/XtD1ziuTQpyB/b+hOPIuEvHRMDtDVtzw18/d+bHr0 Pt4lFPr9DuTvBV4YIAde3ZQLFdpzBqR3PRj5U6XRfsszNUq/GvBqyXx3hGXwQQoGUVB1gZqPToIu XdsAJU4Fg2dPJy4KkB4SDJdwEaVrUQuIqASDSklPLHJX3z3iSxvK72FUqPpkFXZ0BB15gT91Z/2u Vn1oMeoPAm8YMuSS6MjDVaXmcNgX1iqFdUic+Sr+mlyBlPLvIEftMQpGnZpiYnu2N0Fe3OkYJEFf p0S/iTc3+a0S12beNtRGsJg5Kn7g3kBaiAHRtkFtxXUc5SjI2SGZLN6U+McDzCVVcrP98akvsEJ3 PiRRJ2EE2r6iaYdU3a1XgHsJwA+3cA8iQ0QK/3eAiUtICU4ofEP0txR8LBJfhh4YoLrhdARlIjsv GJqTGh7Vs1sfybgN/LE1ntS4oOn4nm2NSe3tIwxUFaWDMidwJBFxfg/CdND6HR5kFblQrTPw0V+0 56YmKFqdllTcvql2ZTVclplfRb4Gy7t4XrnJJLc4sI4chLiuh/zYxu7bjlqggGBjCtBBNVtHnyE7 5LwwpHCdF+lbxu4uyZEFioSRKqlKQcs+kVyuCJ+a4cTebocFhSyLJDTubNVoheSBYWMthMNsIRxi xxnXwdCdz0K7prWZfgisYD2V++B0n+Mvw0mUnVOdg8Qw+NkqWEBRCtjf8KxCxFdH1oATz0FZUyEm BXupQE88wzSGWNiLOfBjs4kzDVg7WNFk4sAfz2mEov+XuATMQs+qAFBSKef7sCIoJ0PLqrfMQo7Y 3AtDVdkul85PUrKJ1G/qVNueO9BMjZ2yVLGoRY/oIiaRf6AoZ53ofJy5HkDLmIznBAFJVKIoFD0V a/q6unvZVa1+U4eLTs3gsGBpc4gzRUjQQd6ogYgny/r8T8SpeKoCmBhy1qO8kdHZF+oLGKBx1ot3 VekeOt5D6VrIdNxVfBdpZjHSB1X2W2bV8eRZsoZTkNQCyUMp5QxT9l220xBY50Efybcgs+Yz5lKW DhqSjFoRVQXApBOpx/kwB13ExomZKSxMYm2MHaGsMkHgnStUzEXwwyLG6KVRboufLrivQFgSmdU+ iRsjzLWHwqeyawqAcb1FgwGCxdKPthZTjj8zZ+N6S4s6W2UJf2h1aqL7cKF4SaxN/YaRsKiraQm2 enM3sIe9wrtqXZSMS+o3bEbxEVGdKM5pLU/+fhVTaG1CjQdqXIMxtKPqBAHw0yuJccdi+70vgBLj 2oV8dQcmAt5hzexBmDMMPItkcaWa1EpnJEQ4tmaqrIdu77pFMovrWzLUfohSqlpJ0uind/Vz1kCs kZ9Ux4e2867PyAqxiAVkMYzmXfIF7TTw2XWKu/bAJ6X2FaTXUS4SPmddn0QGDnGZZXKNfBFyr26j HCJuwzKw5YfRMAOPLMhTXohTNEUkobs7L3nUc4Mab4i2G0g3RgbyBingpCjCVbtK/kpzKaVROmi5 A9eC070A4ENAq90EqLXlAIoXc5Ik8xDZYS4Ujvldy6ifC6x6KoHj8L9SA6a1GSZzi7eePm6G82S1 2pPmnWtHP2397ZCioiFS6+shtYYCBUtqLWutDwBQyoH/zuhSCxC0qNa0z0rVk3cZTn3fmR3E3hra dID9cglCS66/GZ9jqDWCWFECQnpilpJOwDJlv4jkwZ2w5pFBy19Dw0XUP+HpL7wiElYdPmp7SVUh aoj6CBTZq7o6GM8s1CPUhFKrdiLS8oqTkISS+o2snUi6guDAdCo9hRI3qOpiFDck+AGqK3H+sqY8 H0xNq2Yqqut6r/CXlm0MnMAOyXPAmAXbhhiBlsztpNFHsHlmwpPeS781vBzGLXJUYbfDuUM9kyjd VPh5tvurVbK7xQeSEcLLsnTkh0AUPHJntIiufIY31vP6CNrH7efGG6dZSlUAhblefX8MKLNhRm9c 1taJnFVv6tdicMC0k9czvQCQwIvCAREqaYPprrE9eam6c9aBocgqZgcmpX9DspTllKJqLjJ+2sU5 5Pwz4yfBC56xGlLYw6WQ7p/WOkKyLsKrwCFmUu/+vshwrp/iQ2yTbW+TVuntLfsjheSyx+QcHchW h66C0w03PV4qNqEQoO5g4q/o9g/41Mfms+sz1WWcakdKQbetg7e/ITno6m8VeyOSkafm+blU1oSI i74k5D7oq1uOnAc1olo3qBgse14AoHRAltZxo3pa3h1S0Ke4pVhggG7iUzdEIhcdB9GrDt3iRPPY JdrZLWJwn3rhKmgH/RKhV0P3juTiVVAkqxlYJtDgEmjsEOaFcgWnDhJW03RFb1H1XsO5U5PLMgzb caOBpDLMHNTeFVn2o1/Qe0+KRBVIDhQB79lfUQs9lDO/SDMqA4eQKKoL9dUF0IbMqVY6QOeGJ2wV I5GDebPEIkgusVYTJ7nt5F3KSxid5MnrGq0D5MCQqNyUDhNl+HeMVXu9oW9rK9/k8Uev61nS7WO1 K7XySB0rfRO+rS0WJ1s5DXB5no2e8zq+CvHN1JbaMO3r7DIy5HlGDLNQTNRrxPady+N17FzYxbVJ lN0fhp9eFQknaFHq66D3c7RBL0oIyqy089EHhVKlqDaLOq/HGVhOsgDLT5AZz6zxNmWmfJ0Fc5g7 gax34bFmhj/4XYbAHjFBRO3Zc0USDWPWcCKwug/yN6BxJ/6oL2HHjWRq78lcE1QxdpMW6QHxBr5b mnSdwM4iBxZvPosYLak+N5XXvHBJlEoOOcaFBoR1UXEt2m9SJcwdOo3AZS4GbgoXloZQJJKxLUVZ ho6Eopen7nC9Ma67Ee1121GEaod1Ixw+Xnh5vlDLAXi3dQPJSxnFG95rNVbTdOoyREKtmt3bbuLg 2jwvRBASpohk4vQxyhQNq2AOektNCOsWP7Ibjk1nZvnqMszQQ2fdWRkdZ55re+i5Y3qQsyNpRvUb vW3YbGCIVQFEzHSUXJXFNHy3Ygbie8LG1I5DxieaniK2A7Ws0bIy5OY8pYNkKUSm2ia7frBBgMsj GM/oN3OX+fYhjg19rfmxM3XUeh+oSD92lrL4vJ0o71W3ftxaCEsYBhS4Qfp4eNwYZfU+xlDq+B+g Qr5ayeEKef5ssRGO5I9SIV8tqh/6diAVSKm7jv7YOfLRv5JHxWDRRu7iZQGW1hhs3TdkIlQ+dT1L 5VitjNMPkIJW5QNJHEv9hnEsicUU4RIeNdFymoJj0U/CTTUchH72UwKOulZnNkVxgbQwwmOVh75X 2F6BJtIG1McfYAMS2+NzS82nBeZcHXbrwXs1eivyBT7wbl/GO96MoBHCK3ZFu5+4/hxd0yT+3nMy bbZ6s+Ol9iw7XW/6toFHc6AOCXFZ9Myi9IvD16k0QXba5qilOEJNWoXo+C2OCsOWJYXplK2H2sdz AhqvjydHV6tSDjM8pZ4C8A9EWf6c+xBEMggrYNct9AQgRLtdukhYgUy0QMABOjSUDjgqEGnMi0ig FoUfEDrMqnngDZZ1wQXRzwvlXzfRQ20KXh7c+X2FAyWLl/r/wWhJNyyXicV0cUkC2T/7VZ6Qq7HA SdjC1B6l2dJOZkknHLca/WkXCTauZjZDu6H3o97UYsUKlbGv8vhfs0ggYuwjBh6OYX+BwkKekIKn Eh5oUbIqRe+nJ3UwcWI9ZQq5bkL484I5b4d0rwlJj9LNczxIRyzkpYzZjsqBZw+mCNSp2o1vofnZ uHaRz28HlcnHcrm5zmBfZnvmuVEgJalRpNecAP0L44PofnReFm0wX3eJy7oJvRmSIZSLFg5pBSOt /Dl+Ax1mIo+jlpG2Tz+fE4eG4x75zZTeXvcuZjIvRmdvC7Ukknp4aFddq4D4lYSycHyBOehgejwM eoXlkQCG9BirFuS2cBHJDLW6IKpOSfArsQNAKb/Aw4lAK0sl1oFYGBvFnu9ngXScPTqsffrlmW4V xx90+1QSCUPVIL0P3lNm8MKZinaBN4ixF3AhwURCY5Vcw8/7bQE1BXpsslkmUFGopxpkZ+V6hQtW ilMy3fpAd5vKq3ovMcM5EE+AgQ6NRQrz/WHhOWK7LblEobDB8Z1s1KNrF0mmNUYG5qAmklqdAzWR 9Jg01bHWAXU+mNZu3qNW4bOpXXktOp0D7tAPptVzJ26g7GZfXabz//73//y//+d/8aY1TT4LGq0e /evti8nbwjFE7hLaajPZtHsSjNaWpimm1UzUaax/BDHpkNL1ginuXpRkSffqOhweehB0roy0YN3K pHZIglYffmWjin3Ul9ILIg9/Xkarh4AW06Fia2EoiKLXXk493fLi1x8Fgm3r6ADeLSQIiZuqk1n9 ZZxweYUhOraGgWgW1i7qEBpkfbXro7DzG5GXRateHnNcj4B65I2pr1ze/3Tsj32ed/FoFNkOV5Zy +oLyTw/0ff1eEBwhu4SHpW8TtHwAGlX505TrqAMjuuMUaYbfGYwM14skHH1O2fcCZPt5a9HwVPhp xWdHoEcfPXyUlfSJMBJ6lc/xgDHj8TK1HrO4ezCTahl/fh9f245pwQ9iBmjW/dCLlAWU9izLJWmQ q3zmAaDXc2E4M6iENZfIMafUc54ylItTPRjUVRwL0mI6BKa45PY49rZJ4e1Mr7F8AgHXh/ORddwg /ZQ/LV500LgumCtB5CherV1H3islJjCnzmw8qxVC28MJWjSZVS0IGURCiS+oX7J91G8eIzBqwu8Z KAZEz1jlJ15f0ujgrUMLIZWKjXLHyKkr6OAOP8EweaZNgmnKxqZdqCJ7x1RhEfQqE+orFz0rmaic 58IJkO03G/KRsHTQTz+rToAaa3FdP0RGlaT8n7hMaDfUrwM9OUQLfTLd+gQwjgdUL8rA9cSWP7MO 9EhjWVqDOdp+VolbHGFUopBLgNXHH4BcyAFAB6AmQrjTYTicBbU2NN7EG89cfvl0aXN0+2XE8Qpa qK+VSxrOa4kgJc7bjox0pvOqe4mOCpuOwdoMo9fUeV83hDXw0fxCxs3aEO2LNIg4X2uxXgBMwvpu yQHYnexuGWlZhLSKDknaUHF7rVhWAaT6ZmzkgMQJdadxwmoLFlwL63Rsljk1Wnfc8YJI/S0YThFP ryfRNIY8bllwLevAIpd/tOPjVfJiiUIDdgEYnNO6wRAHHENs0yl37LlwHpwhTfnz7qxoy0HsXjew fHgj/a3Ch9dgnG3SpTb4pEyn9WbIs+nvBZEYSgdseEkRVW+sCNGZz51WStVGPJZWXGjFcl76wqv9 VEupLGpodKP1WDqWyJFEnzMtA7Z9Xfp0/qFlmxOfI9hIS+7O3J24XGXrAnd9yxq0JcFRf1wmQ3k2 yd/a7uiUX3R9G91QdRTTmJctc5Ro0bLHsAs0GKBypN7bYuEflFJDQ9K902NqpO6ghc823aDOAWUH CJjxfHtwAd3ICpDFxYYIE1HDBN1rPRYkCnlr64rQ9CGnoH+6xR0sxnWUrFC18VKZq1338XHHJkK6 1GzvUKZ/4Fo8eNVIsVe/kXUfJiBVbcsZzh3PDWv98Sy669ObkxIpeXhPmJHJ7WIQrmBavjdEzIp9 06qCjcqW8IVkKzLFGlqFhvJQilrZC7ykhQcm0QIKrmaKnCINrMhoLKKHyhnVTIbDJUVsgvoQmjWC +bUUqgCt0+tZRs4HfisBFASlxYNmRFGJk1BDkvpCimMgrUTeCxUfL+oKaesiZjyz+2i4QBDVF6PZ Q1mPWH9H3oCkI+ej59T4LrrBnslLz/uRHdKOXhof6ebo6vps3WwyuNrFT1UhJbI3wfWou+8VBcBF 7oi2aF3eUMcUuhVKrwtfzGn2LO9TsKJVf/qZYWdJLZRe8FrfJECC071WAcpkLRGK5mvbFH+iCXJ+ F6Nd7feAQ5kISpsGyyhEDUerojEI6mWXuIGRKovp3mNc/LxG9sHOEA6Jtx/e7VAUry64nfOj8Y3p z5xGrifPR1D48Hhg0/V1BTVIDi31mwOc3wrsYBLUHQV+6KICbKpyfuHoa5lR9TmxGSWvYBHeF1H+ OcqzURcoWWMPlD5QLKzL5dVt5DKJ2OWTU5baSF8Ad25Kqjcixnmpw1F/iIbjQ1kCPtoCqTPDdbRo GC3Efo/y2R5jT7ezr0ZlRKuI6aoybvrrDsCsyf4OHyflg1yle2rU8q3kNtyCYk4bJl46tROEOeEL 5HdGcjZB0fNCh6lb5zUEaQVGu7z9HqBoP0K9jTULoDHyhhCPhlqjy7wGOFKTSofg04EGGgihDGQ+ qV572CjNe04elehGSbzCDeGsnp83sjFuY7R8F90vqCyd0HGX7jP0tSRZJXW1gTJcNL+hXhMZ86Jn BrXopl85CYjLPwrZTGyFVYmfsy4anWSh2j013TH0vJnHr3t5NJxpx2Xy52KDrLEMtRP5hHYKLHKw Mcp+2qwLN1Pqy8If6X479q2Q9Czk8PZyQoBF9w7Wh/sZHQZjAfNPl2SvVu+eF+4sNgZ7uSf82rUH G4XZY5PdnHjIblS9+Af8+99blrfvzpv49nA2qJVxOnCSm9NAq6kcCmlMJ7iJlcZlLOSp1Jf+fOEe mgq7376Ld3I44Wa9Ma6C6rUcgKwtOqAsUO917yLLluaD91mR0ANOWUtDY1FkjSCkXms7OTJBGgms RyPd56iBIBYOLy+YUdXJKGUspu7HgCbCH42zHBZJ17mTK+CYfXSc+j1grNpGoMM2eDYxDYxCyIsc nbXS/fa4HTaVQ5yQPGarv6IuobTL6kDvlqhK244rdBUm0n0gj9zKxWdq+MUzORnC7CJgE1rQ1v0+ rfBvrOZUfO4AShGeIGFJHvMq8bjdEf74ENMqLYOBOwh4MKGOEo8NqbcpqqAYZZewElW0aPqbf3tO vOE6udmLJj7kDoNHB+9QB7UMrgquBQoFkRq9GlqFWoryPdpWR7cg4BtqkNR++szqEl+y1m2sad2S adzcEUOKVozSMBFKYu2LoIgVaA+ao9cZyteCkoda3Ur775eY5E9nQzjUqX+3pIm5gTMNwzHlNEjG SmdRhmUFuAWxeFj24zDMLwYR/nT+MH3TN2BVDEDs3w993EdfzPVeaAL8Q4nOqjS1tm6tGLDC/Hjz 4o8fzhSGuswiuowc422RmokQLhASTrOPWMAZ+21c/vam/I1yVLCuLeAYbRa4uoluJfEms9nQHVJb VvoCd8bl9LHj2u484K31thfx1xxmjLH4iucDK/BcjLP4hqc8F9eKmLQ2LATXpiA4PePPrthkOfs3 Y/9evTrzh2hwyB9epriIxkgwFVx0GzDPV2fvEt6NyxGPLN5+YV6CZDGHwz4mp1d0wYL74pM36eLz jljgZncRvTq7zfPtxfn5DuVd62j3Em7rDb4D71tHOf7Mbs4F2NarcxsHdU6+RUyeTljsKiRXAxZs cVBvL3Tz33dWaStT6g+4z5ITNrDlEIOGid8u8L9YFn6792ibL+8SdlcQDQ1QiNPCybecVvEMfyOi pfDDaQK3gneWpejdGS13BczVUc7pT2UVaKm2nSP1nYBHvxvZRby+ioFDQGXy/0WEqW92ufiNQ/W/ 2wNcvjq0xy8muOn6BdyxsxdIVQteBCZuRDPdgTWxJv+D3gb+o4EjFh6tptukOGLLbcByndC9nSgV fgkl8jylvuJxgVxALcsUqMWcfIw8ztnSoBNVS8SvtAla6y5bvAfugTLwe57FuHSKfr3GXsXneLj8 ggGmggWBaQfPsHF19zOSFF6dsXvN6f2v19mafmKBBkgXRweaZcthqyjoeWCbKIQt6BlX7DECqF7f Zrv8dZyu4bTfgT4oA4XBOvoCUBNbwnmLR2i2TUpnxPay2igf4EH6hK2fVix+xQbYIIK0iR7kv/F7 yeXo94uS+ZGCyT5pyXpz/OEUrW1q/n/T84bTmTMu2TPgYao+d9mCV/30TG5AgKEgGvsvi+22ysXz k4i1U2Is7OIvCzByhgS8ayfnLEWV3lWafl5H2WfmosWDxAzZVTScwJQMdz4edtsplHokS7uQQqKk piM7WV+BIL1fZrQTlQD94Lt6dYYWfxT3jc9GxofZfzHEiv7l74zwfTiGL4cOr3yRHRzbUCaLT2mO jjW2dQtThrBdzzR5N0OhY1XCVrZJtDnh0jo0g3QssZbkXRw03M/lMZMU5MZZl5JRhD4+JbhfEipX VwqznEjJ5CWBGSyQUwcmkcErJQt0zduS9YihWkjMtcbuxG/0oHCswcSzSdPoo5qJMjfxcF01OyXg VSlOjRAT7UPEzhj4D8XRsKqmTVrqa5qJtE0Q20nmIzRq7clVKX2PZjm0r6zwrdZX1wuEDM+b7gFA Vgu+nKrRJbv1pLnvcXxdxMFkVNtiul5oe/bRzu+Hr44VZE6WiDk1PWfC2XQH8OB0q4NPHkBJ+VxH 2y0CRXmqe+PwlFXaiW4cSQnIR8VlQvUXepHHI6DoD1xYW13eqq2xfXTI/SYLah8dau3SQUuiUrzr KKanlQ0dSpded32UlZTKbAdo3s94tnL5XIdWdszi7sExqmVItkXg/l4926YoNk2TY05JVntOQeaD Ro60mA7p2cMglAbRGDn5yK7LNlkMNwfpl0l0ClhqhpVi2jzKOvo5qU3f8qfTkNwKkpO6tfWRaGtR AJC7h1UjSPZAqI8/hHnUO/GsMJbajb9g7E/smUUeeGnLjuW7KIKsOtUz/8qMbiIs7UEpv1J03RHA YLuTGvEUptohK02mtaN3J4xwgf5XGAv7FX+hnzL6qBUJZ/07D0rFNQel3j3YeQfd66PsrnoYHP/E h700CBnaWhLORyzJF2lRZXHZF+qoIkoDdmVCYBGK0jGZx1lhh9hm8UDdahBPmvu8vO7gAcvw6o2d nV6+shEli8DV47paXexUEN5/0azEsDxuMuypmu6AcfcdQDxCmLPs8ylKj4oGsrrjPybD42GPn3r/ 4OaZu41x9Y1hAK7YMHDNS3rH2gKxRJYUd9ey0nliXlUTyPolyjXH4NPyrJfGZc7izzWwF55T2UXO 2nNyF/kvb2Z/fnP59p/hbWRwbnsc4Unw2NrzLMbXDgFVMvthMPaUnq4PoNzkIzooBB/LTgyN9Jhi FSQ/n/ZokAuwWO2XzZLAAtiVV3uonkL1hQgR11zRR1D4w0vJ8sq6Ba6r2aGYEiWGIBH0pUfTi1hf XS0OpDijIsT/KJgSbxbRdrfn/b9q1JEVKyAsOVUmHZDh57U5G3n1iJqvVl1VBo+0qGXcDZT2BQIX RQhNhtkx/SgfHgX1pV/HUL0sZA6cJwqrKFxdO1UFk46Z+eGhgvyqmCXM65Z4Kl7dX5eh5Cm6pndH 2aOUW7Kj2NwqjSByI8Q5Fk1G+STA7dXbBnWrVhiEtSB5q92qsxsKkpIsO/XxJ7HsOuxW1x7PrXHd VJ+FuK1oXqVtMbk7Q0VfZcrKu1MyzZ5kd5LywuVPszHVAU2FbN0ylnl/wmjUW+jo9ztW7IXiHpgW /fdoe/f+ULoA/9BaaUdD5oTYarvZW/Tmp2xPueFoY7lPALaiUplcBd0+GZbYGjXWW7AnomkB3uIj TjrD0BvOjimaeizS0ZGMHbhodUm2DKvO+zFWXvjGdAt/YpCP3qWZNiX76SiOlcc10PVRyatZYV30 AtEdYYAOeO4jV6K2MyXj80Z4OxreDOrPRPxAFBgncCgVbAIuJvhDkG21vf22Yz2yiyD9S902n85D 1kCCgsyJZ30P6I9Y8w5Rcwuo9gBTsUTOiZ4Qdqv4JlrQPWSoyIBjZYP8VP4Hq6gtcIJQpceuHpf6 CIm1rQxUvfn9U8NWuWzpycHXziLobpnLZ30W96NgLAoQ79LsM+NQPwJadsC1YJ4/KGxH5WXqAp5P Tt2qcGm3P82x6Qcz1glMipvaPqrNnUC1P7vz0tVvnsT+bN+db03nQ9wFDzVV2p05dvy5Z1KsuEoJ rqmFknWtfvMku2tY1/7YqV2S+1jWdVNDIwW2tVPbd449GmGlT1V9FgouUTTlKmMaBmvd0Pia93I4 R4kixHXtGTR86EVD3tQaT2a1q51b3VZ2pw9H/eZJsKydhryZDRZh1pxynjO10Iu96i5NmqHHbxgr yEqiobHlBAO14lE8zBSc4g2uX0wca2oqHQQfDbkLg3FeNX9+h6LpHwK9P0JD3MX/tkfIBQ1drsvQ 8HkRExZ9SKiDC+E/NzF4dzw4TVivGBLnpOQVIj7pr0EJA/9RFdB2SY+gclHm3BE5bmhXT7BKlpVR 5j9QJ16dwJ855mD4JNcHPUm5RPvBqjUUvPbfOKN0DKXhMz3GWjGfIZDB75RixhTZT2XvPepZRP14 cNszOhmiC58O+iEq+tWepAc4yoN5MYmIiQAlBsOKomARUrSGNy5F8skiS64oMwn5H8hFR8BplycL FssBCaMdLoPS7uWZdpvi6kwiho5krI9ZhIpqcIcDDbtbL+F0xoEV8r71HcOv4/w2RZcq3VkcXiRa KZV0VXIuQOsuyijuz/Gm1ySqXJHXj1wbSiOILyodosQx0eSrmvF+2yndQKXugdqQVYLYnL4/8Yln oA/jytAfUU+sF7yj2VWWLOk60qLF5XPj8h0uvriC4NiVSQgFV+aXf/WASjf82xnEcwiyVQLnCV8U KWVYE9p3UZ8L3nlvYYxfvyvlA1/JS+P/s3dtvW0bWfivEPuwmwJuI14lbbECKJJKvUhaoy7QZ9mW E7Wy5NXFqffX73fmQs4MZ0a0JNtptg8NUEsi53Lul+9clj/S3wDJzKqNuFFH2xH4XcZa7Ul5V9XA 3osY/YTe/kkb2Mj+GnvuX70Y++F8A9hpBfOYgo0QDEig1xwt2O8suNoBHQFfni42KwGasXjswpVP vTAOFm4+ubVvtKlpCMfyNSyc2IviyCtVpPFiXKKWx95/R052tq7We0ly+dxcVcWK/eakxEKFVYc9 MPE9qiFoz4KbR5SL+Wsk9m7fsTBRPbBbzmFnBp+nj97KL5VGjYS+4qEqnoBHE+knYXc9sl41xhBc cyx2iMxTVOrBCd3BUEvW9U++IMcqTgfjfjmhHJoSnEiqZFgVuR6cMBwrEYbVsv7qlvNJmJb1E15l ywo98NWKNXU31A9Oao5kZ/oHPk0+qJaATV8t74AZ42O/J1csnMxI/CewyjHx22r7Qa3d3SM2cYZC pNUaIOhMGQqRSBpu1uxO5yk0l6/3IM1LQUYs+3rbh2QmUE2p3313lIRpkmljvV7KkH+zIZi+W3/w +SXOkDrRfUeUp1n4Or4OI1Ugy/vZ7CWucGSqMSsn2HX868U9qJpemtiOwIcejlSELHHwS5ysw47w EeTrTU7mIyDQtl+j47FGBNX1bs/2A6k8zG8QCyanw/VV2N51I5aJ/E1lmKwSgOIYk8ZtZX6/foF2 s6c3TtLhcEAKXDEMemlvGI77TckNXblR+q1EXHXcN2YDKIW0ei9bW5GhFg9rYwF9WqNoeZPqgmt0 8YbONp5Cra0HiNRBA4yztxLwWVHoonGZDeMhBaxVFDqa7jsBmAzBLRkodFHWT5AAYKhVgyQLhxJY bj8KncSrayDoYPDyBpELAUGXsGJxPwQcA6X7C4LuC4CgY7dFMGgEU/eKEHRocyG8s78w6J6MQRdl mOLdcHMUZ8SAwIuTEHYSYO5PhUEXleWwiBHB1fRKUmBe1zhtGqJJr2RljNmx+KJa3SkCHfKPirIR ve7iE6Zs0ACUVnzOnlAfJwWmC5n7z6Wkhh7BJKUXyawDEIGCtnYQuofy+xrkwmE5/RmQ6bx4C8pe W1pc2hD1LHZQiDjQXx7vAdE3431Q9D2yI9o9fLhn1hfAPnZkHYILWVZ5HDYdWTq0invGH5iLNSXo 2TCL4zKZGDUk1uz+cJIBgLbmGpVBMIipwRxgDCK+zF7I7SudQX4AFCn8+0js34E6gMABYdUC+Xm1 rjCTMN/Mp/wXwqhr+q9+m3777wv6TN4K7Vc8llbhKt+wvgIX2QmKrvVSwgIUAZr21DgpWkgEKefz civj+bhfpuuPMz1WZKeNtOyX/dScsRUj15tOmJxt6ouiMUY/Tmy0MYniYQNRwWhDCSnLG+IXNWfM 0NVeV261ww+dSUMCJfejOBib4+a9sgnzAm2kQyuUlElE8qYUWVGMv23PL1yrpBKm2aCo6rN1vd54 g0Mg/yiqIWUhuhlK0N78DBt3LGs5Q0gAmVHMIEI0c7sDHvx/ZzffoQiKAoTr2afZcjN/mAULZI5p UBX+dDtbT68W/trpQRLGdaT4Z9fJ1UG2Lhf3nS8YcNKrOr9BrGl++4hW/d+oB3ILlx054qba2nt3 4yEGozKAW6WDpU203aiGggUICszXrTAuUtmU0WeRB3lzGyZnvAeVPMPN6LPj7SItSfI8LGG5aUZi mkVhGleNRcjiTT1YlHVRlyAeTC1W/6jowHGYjmMCdmdCkck58WVFBwpJcYCE8/7EKdug3PZGzVst 5Ypc3ffWkX7lBmu0wpLsVU1c8gkvcoiNLVNlaKD7zw6z2mlqDLUJL5CrptwB2tanV3NMlkC3irE0 TczlVZQh4Xb8cpABpYgbqtWooo3NHXunF9vZqTLr98u8mBhUGcbjahjmjcVloUqFACtMI4vrikWH otWNsPeQphcwBBnOKrerlrs7oYsXDwsQM2vCFIAf+OycMJXZ3yJmumE/4he0M4+51Rat+EHNDpvt z7CVZ8BXuUAOaIwZ3b8zU8BhEuftTDNd3hZlL4hOirKjO6DeLMxrV+7YI/WbpTkWAEJDP5T5cFVp C3HAlmUn3tmS1BdV1kmd3Aq1QiGa7zjRBmRNVuv56h6KojcoeTrKcQ6WBaoPyIs4BVV7DoGKxzTU nNaCPD8mbjfPR339/vVbkpLmAz3v57hKpBs55hLjedvvBZ0fQ3EUbbc92nExZmC+9Vv1nFRaJZQm FOgxoAXb+06wFSaXr9m8n3vMdyWGRbHYGaT4BsIa9cB8EBBNJUbRWktyey4EWeTZNYZ70CghoRNu pD0pM81vHcll8X1Ko7eL7+idJ9g5kcoKVEvDfWjICg560y7t9GxQv0YSufRl3aPvV2E86Wdk9ir5 lbBXpYMUfdfqxQv+lH9SlIn+CVMmCjNbQ15fiDJxMMTrqwwbNzkWq6btSMgR81/IxJ3tOSegTabR pFqwc8B2BJOGe2dBwes2zMVIUiJTJa/6NKCbkfPorVBy5g98xK7WfsDnIx+Re4mAH0f+Eo7QzRzJ TDiL7IREMeT8jspI1gCWIhtQVlCTcymMVSkJXDUnuuPiYjKMZcIMcp3J0l4y6RVl46uzY+BKUJ4M d0KLsN9jCKY8hPLFcp6DQHP4pPf3dL63a4yqIRqFJn/HzaI6Hc3t8Q26wa+pKJulnGUFAA9FQWM2 ZcMc6kdeUlPkKrvJdeEHgSyPlE55MIAJ7a2zJMXG13neehKnUcYCNGpYqbV9mE+1IlSIcAIkoopc UY/ahVyQK+jHIUYFazI5RrP5ZIjsi7oRXfKqdW8F2pQq3cNU/PmXlsnbEURqy97mFDAlDwyWrcns 2j57YTbmY+QdROa2i83niutbiSqrR/PzJ75XtmWQuq7JU2t4v1+RZdIiI/U94SAtk7HP9lV75cE9 54w/ui69aVwU9kzdCYiCdhZSZAiPiKfZFg5hyO2Q6Q1i6/ACaeIiwms3K5rS1nShLPm3eNfVav0N dRASo9uNshEk6g5zDBkfgddQv01l2rSGu/sFIvnADACAJk4VFh+EtMTN/KnzpnVeM0TAuIe6Eo6b 4aIo742J38tL5IJ6/0Pfh2fB+4i2RYLhfdyQjPiTsTt7VMRVnb93UU0IZnbT6U1N/EXulOTn3vfY nVjs/C22bLepW1SivlCzDvR7tavcuEyrYS8/LlOl57CYXatkYqwy9OvNVNmvtFixVn/ZlykMPgrH 7xVQIqpL4ti0SGWi7hdbRg900Y4P1TVS8qf/+pst3WbfxI+r5bfCkmArNzhj/aJLHDlsX5UdxmjZ 5wNAn3Z0asbskIMKZCLodU9IFxV2AZBNykkxKQmRWHFsM9SmZWGsFw6GSRL1vtZ0pLG5k2S13pjX r9JmxxcaqaSRt4674zM9ITNMN9RzqnZJsLreodF0iQYxeCDbT/Ajd8wuwT9kiXxGtIhPlZ5vNjuE Y9m3/HHFjms3lgcY6lqwnuC0jz+Z4PN8+ymYw45dNtLSt7LnSNqRdPa9s+NZn+A0hA1zFgDxfYYh 3Zxo5nz69gmWeCA59IHmUjRhOxe3G0+384K/efaw5DUCxghj1rgMZ8HdHOkk8vd9Z3bYruBgsLch 40zxBOlASR/trdD4vhe/HD1B0rBO6GtK81D4iaIPUvb4lhj3wnhAgCSW6kTFbjFu3N+x+Bys2yXf DliUNI/Z6HhFZ1vLy4ybUUJi+idqZpNZP3r4Oxwmg2pcGpE56xtzGA5ot5IHrbxR/0R9I12AcBOE ievNkZMlc2BWdDsqXFTChYBYJK3I4e82Ckcwjo1N20LB80RSmd1XxWXDgdza2qTpUrRdBpy2+JLj QFprV/jJe4/OI25FX3yv8GgpZekji+RwFxM+yzG5xNTRxwX5zeRgU6SBeOt006lIAkXbg3BoNBT3 4rzsZ8Om0ImF3HXG1o6oG7crvGUNChyV7BqIMlV75cR25MSs1DbCuHREypchx7i4ks6jSOMwLHyx SBTUsb4wGZJ7whLspkatnFkVTIsp5C3Q6vpJGA04IJaDbTtnsy0HRBJr2sTkTH41MNHkxzz7cOSZ 6g2PtBAj7tkl7IXEbZznPSpEUzRoPCj7IfroauVlIXtFn00q3H/tIvN0LucRoqIviMIt27DcaU30 ksrOKGVzO/+445Yn8+xEEie4h2AXUXCm/lhFGDBYUAg7mxG+kjQfBeAY+/G0ThbupwTYsmnmHa0c gBJMBmOvOYAeEiTy415OSSOVHpBNzNKiCXnQQeq3rqaS9E9enx6qD5c+Xuv30wkKGpjEs4sbhxEz +uny6AeDbymwb67Ps5jx5SX3UzikF0x/SWLSQ6mTNTLVyKjBfIUmJLN+UYrkxoiTzT7Eh/2nFkx3 H0VWBhPOpkv81wyNVnN7sz9QKbVEWl3kPH0rxWiPFAa/53xkJv/Km6jrsHzfMlSupMmS3iST+i7D kbLHIvv5MMrziZHQ7WHUySQf6vn/pAjzqPmTIpXFEsU1vz4XXoraVdBCnY+kk2Mt33e+wxZ7ZHc+ kv4/rw1Q5KtUrre7JXfgwRmNoG19+gZc9zC/RkcAyqSmurcJQ99HYNpa7TcYF2W/l06o5luRo1E8 yON8opuT43EyYbNXLQ655FH/tT5rnn6fOem91rl2VDhWuSGmiSF0Em9RBWli4d+hkE9Ks7eNgHKp X/q2lIoyfVjjIfrWtF+8KKNVEIKB1VW/h+OI+p6eKzseiYIiwpGrUeJ8P9akyBnViPi+rb7KrtR0 o2ajc4CdqqNerwS2gmEdZFlUDZUOvZ/pcvVUqGodRFUEsA1pWjK5JFiAeM5qLepJUhRmgYwWs9st XkTPuZ2vN9v38yV6+UQtuenMy9yVNT/YMclFR+IJtnhfgUW24wkdU5Aj1OYhxncXlL77VvLMT9r8 UStjTWb3FC19e87yG9oK7TSEHhkqXTEiaGkBUVAZHkeSZXnZCEtFCHZq+9O7EVBSSMbIL+jo570I 4pgo0FnfhDIUSsn+2G++ww+dcZ39XYHG3rmbogh685Jt5GXGcM00XKtWBKF1J9RgxwWZ7/xBIwlD A3R8piew1UrTwT74FCyQYTGbM6ybtVerPMOyHCJ4O7v+tGRDbWABA30TRjDKGlmjImxllKxvZjdn we1sut3hf8gPRf3g1W6+YFiuV4vV9e8S57ZT8P85AvV6mMG44cOyPo7T8pHSIA/TgoSKxYoiBj+A X+yrOOfYt2gTZQldluWEH/gWbpmSnQpm2+t27b1cHSnICpDvedMTeBIG//xFnBA6EJD9Jphelpzy H8MpCeQc9b4rDACGtmx77OrZF5OwGDeKxXX2hiyz0wOvRPQd/Cm3iIHhuwXCqPCVccK8Xp0yf773 PwNrjH5953vjKXdsFeaaH3HYbRqOuCG3DsuS2inEvByrJnJVaB62kBFi2sHsjykVxqLkgHQixkDO eb/qoymxrQuyq8Y06feUaqijOOeoczlsIaOLBXDwyQ5smrn+j0/jAyuU5seBMpCLzRFn8QxSxsFO FadraoJmBeGsXB4VNUhsQjx+pr6B4O+L7fe/vvv7x+33kJVoFzA3pgmQ57CKdHfakC6oKRxjooti lxzjTAQVQMxX639sAu5W5Au0VmMmGTXQfAQM/D14H3AWVHK7Ce52gLiA+ljP7lYP0B/36zlEBQWR d1eksMmZ0w/L4cZFyBAV0RgWlxLgshZCiPygVMCKG6d/wkIBiotjDQW43Dhvfl06cPatpFVVhhPM xNS2kmbVcDg2uq0AV1MW1poO/ZPn3wqCEfsAihXy8p6O00tFtIQQOaRfAn4jU8OGzBm8wV/f+fF9 lV4HYYzvW9VIFgoE1DZ1s0YPyZII9m66nqPiEoFkWo+aSWBd+SwqR4WHuw316kEaEJa1z16Jsl6M QWdHH9jsgXjOYo+17qoFiyG5g1yDpy+nadOw1cbZ325X8E9/t0NKC/AMlpvEuax2G1yZ7AA1izJb K6RIxKkWOPouOBfFuZS1RQAIQzlAOFqVSAOh3VB8l7qRfhEOhnlGqVFFDibhYBznSd1Ox0Ki4mjl TStysEp7GWEmcueVCQ8FGa6bHOwOrTHUC0Rq39jLjlKE4ttPYJPRJQMmfyS2mG63mNMDNHYWseDw JT6mDHsJgGSPZ0rTzFzLC2BeeBqFPd5F3FUogaJlB7W5/CecjINrKFHq7kzU+t8aDAJzGUduEHbB GkLffOrxm+N4DOZz1dUWeTpICHm1G1FCc1G2muAdCI2Hoz5QTqYBeyCrkMQ/IoI3ul2DV6ivrp5M CY4r/IBMAKJ1ZFXLhDZrDwfo2M2UdQPwYdfUIkCBErE42VDJ+9D1pdrtligpkQ8OjdqdaDgs8qJs IkuczFUBo2ZjxImLk/j6RY9kXbIeePEYdA/HLG+4a4ZLITsHJjWr8yGxRSXhMJNJcQgGCTR8EUmA zPpQ6K+xmt69Ki1C580wF01MhKKaEyQ7iCJpY6Lw4nnoNJyk/UnJlKGiIqNxNkjSQV02xlTkeBIO FUg9RUXqn3xFdGqXIpc1rAGDQJgGV+vVFHO82BwNmdFmWnWBHPkSt/gAvdpQ8PT6ena/Zb0BAFGf XzPgGxKLuwUMYxKRso1aqhEmr0SVBD1Y0jPGijUwOiZxQ7ga4rxlybUs3W7C/X8CAAAA//+8VkFz mzoQ/isaTu2MGwtjwLg1MwRM6kM7ntivPXTeZAgIWxNAjJDtuL/+rQQmxnFTcnk5JGF3tdpPu/vt ul8OU+FG8ZbQPUm+DOWX/M1dUJScsXTOOQLpsSQzrSpJls2LRBtKtbLhS/Uni4oNmO2jbKaR4tM/ K2ki/dRq4d50fIOmVDeggz6elhGPFslM0wMH+yPnVlNSQZ6FlJqTkW3gYAzSKa9ocj/TML4Ndcfw WtGSXxEGJI12mXitWUrR2LNsS9dUFHWQ5UocM3JC4bM8J4VYQxQ13GKX13Y022cnK9zqFslJNnIa 8O2JPo8kn6T2/6539SAnsagQLfYs21PIQg6YaQlAoiSnBa0EjwQkFyUsj2ghLZEgPK8QSxEridSy YgBikB4YfwIfAxQVCcrpplYiwYYpZzkqt8eKxlGGCiKkKUp3RSyPI5rDjfK9lLeqT7ZNc24ZE8/p ZttuftrUylyNLGz4Zis6S2xXoxJr3Nq6N3lHYt+TneuFO7Ydywr0URcKtm3fmQRBG7eEYgSWF0rD upbPoHQ1Corjmc7c+zOUryRKIF2jpt7O60e49yQlnBQxqXq0ngU3WZ4ddhFcTYZn24b10o116znY Dk15+hWsrrmC1YhaDhHur06Ih+kjY095xJ9WIuICnFLJDurViigHIlqs7h7WUMS0YBnbHOsmvE5X ykVtoBpLuHD44jp+irsDRsYn3PN7Lo6dogRGfIlRUeP1UF6Y8437/kUBj1KB5uvVAt2t0PfwB8JY H6DvTc+FTc9V6AflYhdltFJNhz6A6ceLEF8FAi2dnz/JG5F87uOrwXS9L7Bh2paD8f9eVX9GhZ7z bFqVUQxVVHJSEb4nmrsm8bZQzLYShHBJo5L/qYAv9GG98j9+Rmf1hlLG0TdgU/S3J1ITs+d7+wya tZR0GmW9/Papp+t40bwQVFCipgGUzQB5JacZGmHduLi6bprrbgaX+LewJvCMFk+Iq57li0TXJSts YQ4xfoQuVmx20XDhyDKDsNY0M7Azi7+e3DZE124VWyHK6XCYsPiRPd8QUdEbxjdD6PAhoBpC4oY2 /hTce+FafkAfDZc/Fw+gg68H3cD6wx5WgRtw0IEN5dxCUXvMlW0FT8LQ9H2jR3F3qf2+ZplmTp1K 9e1J0BjX5NJZmmzz1gznr8bo3PcntvdC03L2mLe6bsrtqibpv4XRNT+frSqMCvaOZZvInv5XcEia zkfjsYNVxrcwxAhvpxVEVy+bSb2+aWelpAbdYZoyBrtKrxNGMxo2q9/g+AD1pzvYkk+whf+tiTE5 GXyLJBbBSjlpxrZ6JbrZQrS6bqh3fGRCsHymOY6aRBlJz5Q1CthWbVkQTYgzzcZyv5ludsAkcket b4tZVoG0oSEbN0FAFd5xKmcJtBBZUhFDkIalDkFF1g+uyvGRJUf1DxzZybXL/Q8AAP//AwBQSwME FAAGAAgAAAAhAD4oCXHlBQAAtBAAABAAAAB3b3JkL2hlYWRlcjEueG1sxFhbb9s2FH4fsP8g6GHY HhzTjuQ4Wp0i8aUNkLZGk/YpQEFLtKVGEgmStpIO++/7SEqyE3dpmgxYgNjkIXnOd+4HfvX6tsi9 DZMq4+XI7x0Q32NlzJOsXI38T1ezztD3lKZlQnNespF/x5T/+uTXX15VUZpID69LFVUiHvmp1iLq dlWcsoKqgyKLJVd8qQ9iXnT5cpnFrFtxmXT7pEfsSkgeM6UgakzLDVV+za7gT+NW0Lhh3Cdk2C1o VrY89hFxwUrgXXJZUK0OuFzhhbxZiw4QCqqzRZZn+g74yKBlsxn5a1lGtVadVivzJgKAaFPkzWXA /ve7zgKR+2peyD1FvwPSPZnweF2wUlt4XclyAOalSjOxtdtzucEeaQPpUYV3lK1EL9iT15rnKU6f SFrB943gSuyx+44xEveoyJ0dTEBtw+ghxx55gkcMixbDUyDcl9kg2Q2+6nmm2UZSJZB9L0moN5Kv RauVyF7G7by8aXmZIvATyMjApvquauqnGOzVisuUCuZ7RRydr0ou6SIHIljcMxHpn6AwCQ/bSFBJ z5ORH05Pwyk5RmEzVM1utaEG/dmgdxaGoEYofsnHkU9Ql8Jhb9aS5khPQibHwWF41hInbEnXud6/ PrekWdAfnjoUc2nAfI0hYUPzkS+zVar9riFKd5bZTcxzLptLBH+zmbnVra/hW7jrll/9tORzyfnS MshpuWres7Lz6dKAZVTpU5XRke9IOxyrqM4iQBERLeMU8pNM6Suo4NvVWbu6QFfoBYekPoCdmq3K CpGzOYc/cegq0oa9ZUbNkd8Pe4Nw2A9wtGBpViYoYHjrezmPbxg8gIOc3vG1Pi/HLId9cEbznFcf 0I5yimoAgvFpjdB4jYRjchqMx/UBSzLnzNOj6WQQWLuLqMXl3Vopd+bTGEpEgqvMlM23LdyZ5MXI F3TFjNPsjQ/LpWL6JAzI4TCEG3aJzdZxaV7Y3ecn8OwPjodHj/L8bHkiSFHqvRgK9ILAxITvxdAj QHRh7ZRhyyWL9dRdhf1wB3bHpwlbWH2rdiWpuDJu8ewK3Ef+guv0MkvQzq1EczDn+d2Kl54xq/OQ PUL7l3rflnlWsiu+pfdJEB7W0LZnfcQORghg/9F5o9b2LZRofGe8sAPRZIcjWLUszITHc+lliBLE TUkLVIV5Fuu1ZB4IUDSGXcbRddNIlYepxrtkWqMJqeuNZPHN9bs7rz2/JqQzvbo8v77gK/4bLcSf 5wXiRF0bomeIXxAqGGu+mImm1zskB1/FqrZA/H7zBiZNs3gmgcUkMI1WO5QLpIGqBx4k6YOS+uPO 5/pNyccpsp+dKoFYMKrbPBfRY/JfKnVHlQnV1FvL7BkKCOcchB9WEf5rWFi9mFu5mWeIBscapqgD AwH1/waG0dTAceCo0drFwb4fvZYkJa9SRhPUWefe+1y6ZntP4UWeiVmW58ZPZu3JiBULk9HyPEEu xMhojfQQMis17E8jVIQLpeuV8+df6GOEHPfPOuOQjDsBOZp2To+Do84RmR4FJBj2xr3x3+Y1KvRa MehB84nImuB66ghUtyKMc+j0Lqhtt7TloGuhIdu3ELE0OhmsSsYfEfWIeKy1ZDpOzXIJ1Ws6LrcH 1k5b0xirKTRWb1G94wmsQdeaW2PcLmVh+GDu3lY3C8eieKwqQ17zWkil3zBeeGYBywOo5U43sLRJ Uni/vmKEldx4zKqSl/cIuGgoFr4BXC+B3zKpa4pJxN091qiQrreb9bblm7kCBTMyI4edfNyY0x+E 4YQAYhVpurBVxN3DeM2XU2mu6zsBQ0FicWl6gi11NTM3oag0aQaROGdUok9HdrypzYut8c7In5xN w1ldrJppSJ+8n33+3WBFJDYovyt9Wib/ueze4R8PRDsrOcXkjJdaAT4mGXizGa4cihdq/elySkjw iPBn8K+i/aGw9qlTSJ/I/kORUG8nIAbTMDwLbUA8ZVytIvUNDGzu9vs/aZjxYHK8Fw47UYhoQMbj E785nPwDAAD//wMAUEsDBBQABgAIAAAAIQCNHUijgAIAAKIIAAAQAAAAd29yZC9mb290ZXIxLnht bKyWUW+iQBDH3y+577DZx0sU0J5aIjRe1aYPvTOtfbuXFRYlhd3N7irtt+8sCK0iF6znizAwM7/5 z+yE8c1rmqAdlSrmzMNO18aIsoCHMVt7+Hk574wwUpqwkCScUQ+/UYVv/O/fxpkbaYnAmyk3E4GH N1oL17JUsKEpUd00DiRXPNLdgKcWj6I4oFbGZWj1bMfOr4TkAVUKUt0StiMK78OlvF20lARl4J5t j6yUxKyKUSfigjLgjbhMiVZdLtfgIV+2ogOEguh4FSexfgM+e1CF2Xl4K5m7r6pTVWV8XABwd2lS vgzYze8WCrjFX+kha4WegCxcpjzYppTpHM+SNAFgztQmFh+6fTUa6LEpkf5Z8KdiM+Fc1fJV8rRp +lSSDHpfJs5ELdwJMcLCKU0KHcxAfYzRcUTHbtERE6JiaINwmLMk+Tx82dek+ZikTMDpu+RA3Um+ FVVVIr4s2j17qWKZJXAGmT3Ij/rn0tRZAWq74mlDBMUoDdz7NeOSrBIgAsWRmUjsw2ISCG5dQSS5 Dz08nM579u2sh3Orpq86tw6LH1hdWH7ho4dt2Es/R868Mi3geMJe6fevr6eVcUojsk10/fWFMfV/ DZ3JKKeQBkWTleWPLchh7mSZ7CCyeRIl4e2GmOf7q+WbgLJWdA0rrVWEmCktl1CdWaKuEiQAfyGp onJHsY8Wk7sZQujvD/Qwe7ybzf88PkyWyKBVnheAKmrk1vQkqxHmajIYDpxCmEUuBuMLyXm0L66w ad8xRPoCEsrCkxA1ybVvNeU6Aj67OUf+lcANrfn9/GC683R+e44SNYD+t+YMWgommzvcgHjQNZgA kU8AfGH47wAAAP//AwBQSwMEFAAGAAgAAAAhACFApIEKAgAAiQYAABIAAAB3b3JkL2Zvb3Rub3Rl cy54bWy0VFFvmzAQfp/U/4D8ngBpEyIUUq2LMuVtWrcf4BoTrGKfZZvQ/PsdEEhWWJW2Gg+Aznff fd/57lb3L7LwDtxYASoh4TQgHlcMUqH2Cfn9aztZEs86qlJagOIJOXJL7tc3X1ZVnAE4BY5bDzGU jSvNEpI7p2PftyznktqpFMyAhcxNGUgfskww7ldgUn8WhEHzpw0wbi0m/EbVgVpygpNwHZqkrAOe BcHSl1SoHmPICDRXyDcDI6mzUzB7jDDPpZ4gQ02deBKFcEfkFyx6mENCSqPik6pJr6qOiZFAfJBF 54y0/+3bViBuP12EGQgdIdmGbICVkivX0PMNL5AwKJsLfa7bR9GwHnlH6U3BF2IrHd4N8vXluebS N4ZWePdd4koP4EaKkbZBsmjrUDfUuY1eI4bBFTdSQ/QcrqHwd86OyWXzVR8rzbmTKo0z+JmB+m6g 1L0qLT6HtlPPPVa9Ct7BLFg0o34pzb4LYLArHnOqOfEki3d7BYY+FcgIK+7VHUnWF+vJq2J31Hhq uaaGOjAETSJNyCRsHLWHcXF9tkPjItxG0XY2Rye0Ov7iamt0eupQ3JXpz4QEuMDmy3DbmzY8o2Xh hic/atPtQxR+XTYJTU2vZ+OvVz6CNm/dvLvFOqqCgXJClc3cP75WFIwICu8eFptojnv8/woaJfaW OFTdKbXrPwAAAP//AwBQSwMEFAAGAAgAAAAhAIE4xtgLAgAAgwYAABEAAAB3b3JkL2VuZG5vdGVz LnhtbLRVbW+bMBD+Pmn/Afl7gsm6gqyQKl2UKd+mdfsBrjHBKn6RbULz73dAIFlhVdpqfDDofPfc 85zPx/LuWZbBgVsntEpRNMco4IrpTKh9in7/2s4SFDhPVUZLrXiKjtyhu9XnT8uacJUp7bkLAEI5 UhuWosJ7Q8LQsYJL6uZSMKudzv2caRnqPBeMh7W2WbjAEW6/jNWMOwf5vlF1oA6d4KS+Dk1S1gMv ME5CSYUaMMaMtOEK+ObaSurdXNs9RNinysyAoaFePIpS+CPww7cDzCFFlVXkpGo2qGpiCBAgB1n2 zkD7375dBUj36iPsSOgEyS5ko1klufItvdDyEghr5QphznV7LxrUo+gpvSr4QmxtoptRvqE81xz6 xtIazr5PXJsR3EQxsi5Ill0dmoY6t9FLxAhfcSINxMDhGgp/5+yZXDZf/b7SnDupNnAFP3Khvltd mUGVER9D26mnAauZBG9ghm/bq34pzb0JYDQrHgpqOAokI7u90pY+lsAIKh40HYlW5+kU1MQfDWw6 bqilXlsEJpGlaBa1fiaAMNLs7cAY39yv1+t4C05g9fzZt9a4e5pQmJTZzxRhmF9fk6hx7EwbntOq 9OOdH43py30crZM2oW3YDWzC1TIEhHY17Xoaq1MamFZeqKq99A8v9eAJOXgRLZJ4C5Ps/8qZJPaK NJDc/z1WfwAAAP//AwBQSwMEFAAGAAgAAAAhAAVG1Yu6AAAAIgEAABsAAAB3b3JkL19yZWxzL2hl YWRlcjEueG1sLnJlbHOEj8sKwjAQRfeC/xBmb1MVRKSpGxG6Ff2AIZmm0eZBEsX+vQE3CoLLuZd7 DtPsn3ZkD4rJeCdgWdXAyEmvjNMCLufjYgssZXQKR+9IwEQJ9u181pxoxFxGaTAhsUJxScCQc9hx nuRAFlPlA7nS9D5azOWMmgeUN9TEV3W94fGTAe0Xk3VKQOzUEth5CsX8n+373kg6eHm35PIPBTe2 uAsQo6YswJIy+A7X1TWQBt42/Ouz9gUAAP//AwBQSwMECgAAAAAAAAAhANZB0Y6mMAAApjAAABYA AAB3b3JkL21lZGlhL2ltYWdlMy5qcGVn/9j/4AAQSkZJRgABAQEA3ADcAAD/2wBDAAIBAQEBAQIB AQECAgICAgQDAgICAgUEBAMEBgUGBgYFBgYGBwkIBgcJBwYGCAsICQoKCgoKBggLDAsKDAkKCgr/ 2wBDAQICAgICAgUDAwUKBwYHCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoK CgoKCgoKCgoKCgr/wAARCABXARwDASIAAhEBAxEB/8QAHwAAAQUBAQEBAQEAAAAAAAAAAAECAwQF BgcICQoL/8QAtRAAAgEDAwIEAwUFBAQAAAF9AQIDAAQRBRIhMUEGE1FhByJxFDKBkaEII0KxwRVS 0fAkM2JyggkKFhcYGRolJicoKSo0NTY3ODk6Q0RFRkdISUpTVFVWV1hZWmNkZWZnaGlqc3R1dnd4 eXqDhIWGh4iJipKTlJWWl5iZmqKjpKWmp6ipqrKztLW2t7i5usLDxMXGx8jJytLT1NXW19jZ2uHi 4+Tl5ufo6erx8vP09fb3+Pn6/8QAHwEAAwEBAQEBAQEBAQAAAAAAAAECAwQFBgcICQoL/8QAtREA AgECBAQDBAcFBAQAAQJ3AAECAxEEBSExBhJBUQdhcRMiMoEIFEKRobHBCSMzUvAVYnLRChYkNOEl 8RcYGRomJygpKjU2Nzg5OkNERUZHSElKU1RVVldYWVpjZGVmZ2hpanN0dXZ3eHl6goOEhYaHiImK kpOUlZaXmJmaoqOkpaanqKmqsrO0tba3uLm6wsPExcbHyMnK0tPU1dbX2Nna4uPk5ebn6Onq8vP0 9fb3+Pn6/9oADAMBAAIRAxEAPwD9/KKq6zrGm6DpF1rmrXaQWtnbPPczScLHGilmY+wAJr5Gf/gu 3/wTJico/wC0CmVODjSbj/4iuvC4DHY5N4elKdt+VN2+45MVj8FgmlXqRhfbmaX5n2HRkDqa+Oz/ AMF3f+CZA6/H9f8AwUXP/wARXQfDn/gsd/wT8+K+r3Wh+B/jUt1PY6Xc6jd50ydFitreMySyMxXA AUE+/SumWSZzCLlLDzSW/us5oZ1lM5JRrwbf95H1JkHoaMgdTXx23/Bd7/gmMpwf2gV/8FFz/wDE VJa/8F2P+CY1y5T/AIaKgiwM5l0q5AP/AI5T/sLO7X+rVP8AwBi/tzJ7/wC8Q/8AAkfYGR6ijI9R XyfpX/Bbj/gmprWoW2k6X+0Zaz3N3cJBbQR6ZdFpJGICqB5fUkgfjXs37Q37Xn7OP7K3h9fEPx7+ LGleHklUm3t7qbNxPj/nnEuXb6gY9SK56mW5lRqxp1KMlKWycWm7dlbU3jmeXzpSqRrRcY7vmVlf u+h6VketGR6ivhG9/wCDh/8A4JxWmqHTl8T+KZU3lTcxeGXMfXrnfkj3xX05+zF+2N+zr+2B4Ul8 WfAH4lWWtxWxVb61QlLm0Y9BJE2GX69D2NViMqzTB0/aV6Mox7tNIxwmd5TjqvssPXjKXZSTZ6lk dc0ZHrXgX7RP/BTj9in9lH4if8Kp+O3xqtNE14WMV49g9rNKywybtjEopAztJx1xg966v9mD9s39 nb9sjRtU8Rfs6+Pk8QWOjXSW2o3MVrLGsUrrvVPnUZO3njpketZzy/H08OsRKlJU39qztrtrsdEc fgp4h0I1Yua+zdX0303PUqCQOppC6L1NeKftZft4fAr9kvRyfGeu/btbkj3Wnh7TmD3MnoW7Rr/t N+GajC4TFY6uqOHg5zeyW/8AXmPGY7CZfh3XxE1CC3bf9fce1l0UbiwA9Sa8p+O37av7Mn7OTvB8 Vfi5pNleqONKgm8+7Y9h5MeWXPbdgV8J6p+0T+3r+31Z3+v6P4xt/hD8KLQt/aviEXLWw8nuv2ji SViONse0EnBzXFfCRfBOnePx8Nf+CeXwS/4TjxoG/wBN+Kvje1EwtGzhriCKTMcAzyHbLH/a6V9z hOCIU03jat5R1lGDVof9fKj92L8kpS7I/Osd4gVazjHL6Voz+Gc03Kf/AF7pR96fq3GPmfaGrf8A BQPxp4l8PP4u8CfAyTw74ZbmPxf8UdUj0a0kX1jibMsmRyAB83as/wCHX7TXxn+M0o/4VkdW8Wwb uNT8OaGulaOp9Vvb4M869jsi7da8U+IHg79m79lLWIfHf7avjrUPjj8XJlDWfhm5lNzbWkrYIRLc 5RFyeAy4/up0NZv7Rf7QfjzR/h3F8Vf2/vjNN8KfA19CW8PfCHwRN5Osa1F2jkZfnVCMA5woBGVB rqjk+W+zj9WpLllopSTfP/17i/fqf4kqUV1tucSzjOJVpLE1pc63hFxXIv8Ap5Jfu6f+F+2n2ufT 13+0wvhTXx4G8TfFWy1HxG5Jl8IeALGTWdQXsfMlJ2QEcZ3BVFc/8Uf+CgnhD4JXseifErxVp+g6 xdkR2fhq5vDrHiG6J+6q2FiMRuewcjP96vmL9mj4X/tsft3eGksvgf4XX9mH4D3JHly6JCV8R+Io enmNcNiX5h/y0yOO7qRX3P8Asp/8E+/2XP2OdP2/B/4a2w1mZf8AiY+LNVAutVvmPLGS5cFwGPJR dqZ/hrxMxp5Ll7cKi55r7MbXX+KS91PvH955yPpctjnmPip83LB9Xzfgpe9L/F+6vuo2OK8E/GX9 tX40xx3fw7+A134d0qcDbrXxIulsGkU9HSwtw8ycdndTn0r1Xwl8KvjPIqXHxI+PF7LKPmFr4c0u G0gX1UtMJnce+VNem0V83WxntNKcIwXkr/nf8LH1NHAqmlzzlJ+b/wArfjcraVp0emWq2kUkzhR9 +eZpGY+pLEmrJGRiiiuN6ndbSxzeueBJ7+5e40rxdrWms53MbO6VwT9JlcAey4rnNV0D48aIjJo2 uaT4ghzu8i9iksZwB0CyRsyk+5AHtXo9FbU684aNJrzX67/czmqYOlU2bi/Jv8tvvR8+eLfjl4k8 Ebrb4gX2seDHk+VrnxLp4utKf0AvrXb5C/7UgX61z3i/4ufGjwZpKeKrq51iLRp4t9t4j8OQxeIt JMfZjEgS5RcZO8NJgc19OXllBfQPaXduksUgKyRyIGVgfUHg14b44/Yy/wCEc1C48dfsoeM7j4d6 7IxludMsl8zRNTfv9osSfLVj08yMK4yetezgMZlrko1YKL81eP36uPq1U87I+dzHLs1hHmozc0uz tL80peilT+b0flFj+1T+0B/YT+NPAkdr4/0GHHnap8OdRjv5LXuWuNOuAs6n/YEmc9q1PhJ/wUx0 Lx5qI0bTU0nX79JDHPosN2dK1hHH8P2O8wsjjukchNeWePLb4e+JfirH4a/aD8PXXwG+MRJOh/Ej whO0Gna0+QCzupCyBvl3LJzyAWxweH/aJsYNL1+2+HP/AAUu+DttepdgR+Hfjf4ItFguHGPlabYN suAMlSOBkgDrX2tPKMpxiUKlBXaulG3M1/NHVRqrvyShJfydD4WpnOd4K9SnXfKnZ81+VP8AlneP NRfbnhUi/wDn41qfe/w5/a1+CHj/AMSDwWnihtE8Q9H8O+JbdrC93E8BY5cCT/gBbNeob0/vj86/ Jb4gaJ+0L+zp4HtdY1680r4+/BG4AFhf6kDcS6fGcDCXAzPZOBgcNtB4wM169+y9+0N401nSV1P9 kT4s3fiO3tYhJe/Cjx9eeZfWsY6izuCcyIOwBwPTvXjY/g6EaH1jB1U4Xtd/Cn2k7Jwl5Tikv5j6 DKuN6s66w2NouNS17L4mu8VdxqR/vQld/wAp+hmR60V478A/2yvh18Zro+F9Vt5/DniaJvLutB1g eXIJBwVQnAbntwfavYQ6noa+LxOFxGEq+zrRcX/Wq7rzR99hsXh8ZSVSjJNf1v2fkxaKAQeRRWB0 HzD/AMFgvjk/wF/4J5fEfxVb3nkXmp6M2j6ec4Jkuz5RAPr5bSH8K/nF/Z1+CPif9pL44+GfgX4O njj1PxPq0dlbTTAlYy2SztjkgAEnvxX65f8AB0h8dDpXwx+Hn7PlheFH1jVJ9X1CEH70UC+XHkf7 7mvyT+BfwJ/aC+NeuXA/Z9+Hmva7f6VEJrh9BgZpLZSdoYleVz0r9x4Bwn1PhmddyUZVG2pPZW0i 3tomm99T8W42xX1viKFFRclTSTS63d2l520P0YX/AINZ/wBoAqC/7Tvg7Pp9hu//AIiuy8Cf8G9X 7THwD+EfxKtfA/xT8L+I/FHjPwxHoGkiN5baO0t5bqJ7uRnkX7zQxlAB/fNfCuk/sNf8FRdW1S20 t/g98TIRczpEZpvtCpHuYDcx3cAZyfYV/SF8CPAKfCr4MeFfhvG0rDQtAtLLdLIWZjHEqnJPJOQa 8XiTO88yeFOLxsK6k7tKEdo2erTejelro9nh/J8ozWdT/ZJ0XFaNyle7utLparfZn84v7bn/AAST /ah/YG+GunfFP47XfhxbDVdbTS7GHTNVE00k7RSS524HyhYmyfUgdxXkf7KP7LnxS/bI+OGlfAD4 O2trJrmrRXEsL3s/lQxJDC8zs787RhMD1ZlHev0R/wCDpL47P4g+Mvw6/Z6sLwmDw5o1xrF/Grf8 vF24jQMPVY4Mj2mPrTf+DXD4G/218bfiD+0Pe2WY9C0GHRdNmdcjzrmQSy49GCQoM+kh9a+ppcQ5 hS4Mea4m3tGm1pZaytHT8X5HzVXIsBV4sWW0L+zTSd3rory6fI4X4b/8EU/2rP2K/H+n/tc/tE6V 4Zm8F/C+RvFOrQ2WuLLJdtYo1zBAE25bzJ44o/o5zxXx/wDG/wCOvxQ/au+NV78UvjT43e61TW9R HnXV3I3k2UTPgJGvOyKMHAUdh3JOf6dv2ivgxof7RXwL8XfAzxRdS29j4q0C502a5g5e3MiFVlUH glW2sAeDtweK/n5/ag/4Iz/t1fs2a3eb/hPeeLNEgkY2+veGIjcxyxgnDNGvzoSP4SM14HDnFEM4 rzqY+cY1krR6Ll12u97790l2Oni/hvEZXTpQwFOU6F7y6vm21strbdrvuez/ABf/AOCAnigfA/T/ AIsfsh/HzTfizezzQrdaVosMSJscZMkcglYYXuG2nnpXbf8ABHz/AIJ//wDBQv8AZO/bU0Px347+ D2p6F4Xu7S4tPEM8l5GYnhZCV3KrHcQ4Ujjg1+cnhL4m/HT9n/xULjwX428S+DtYspMMLG7msp4W 9GUFSPoa/W7/AIJ3/wDBX34ufFL9hz4xaz+0Dq8d/wCJPhj4WkutO8TC3SKS9SRGjh85UAUyLLsG 4AbgckZBY6Z3S4ho5dKm5wrU6nutuNpJTsls7Ozat2OXIKnDeJzWFRU50KtP3klLmi3BNu7aUldJ 6ao/Mn/gql8bz+0R/wAFBPif8SYb3z7NvEb6fpcgbKm1s1W1iI9AViDfVie9ftL/AMG+vwNHwZ/4 JyeHtcu7IRXvjbVLrxDdnbgsshEMB/GCGI/jX8+PhjRNZ+KPxG0/w5ZgvqHiDWIreMgZJlnlCg+/ LV/UL4l1zR/2Mf2OtN0jRbWOJvDfhy10zSLbjBuFiVFH0ByfoDS47Tw+V4PKaC1k0rf4Ukvxd/ke vwZKNbMsXmlbaKb/APAnf8EvxPNP+CiH/BQJvgNp0vwp+D9xDc+MLmIfabwgPHpUbD7xHRpT2U8D qfSvk79nX9kvSvGGj6j+2Z+2rrt43hO2lNysN7OxutfuN3CZJ3FC3GBy2CBgZNXv2af2eta/ao/a HVfGtxPdW7zvqXiS8djueMMCVz2LsQo9AT6V1n/BQL4oTfE34hQfCDwZElt4W8IYtLGxtRiOScDa z4HXGNo9MH1NVgMLSypwyjL3y1ZJSrVVpJR7RfRt6Lsve3Zx5jWqZxz5tj481OLcaNJ6xb7tdbLV 937uyOD1G6+OH/BSv4x6f8IvA9jH4b8D6UR9i0Syi8uy0m0U482RUwHlI4HvwMDJr1z45/HXwT+w /wCDU/ZH/Yr0dG8TzlYta8RRRLLci4bC9QPnnJOAOQmcAZrv9K0yz/YG/YyS80O3jTxx4wjBFw6A vDI6ZDYPaJCMDpvIyK8z/Yw+Gngj4YeFPFn7dXx+nZtI8IWVxeW890dzSTKhaSUbvvSZO1fVn9cV y1cXg69F13D/AGSi+WnTX/L2pteXV6+t93rc3o5fi8PVVKM/9rrLmqVH/wAuqfaPZpa6W7KyPL/i 74s+F/8AwSR+FMXx4+Odtb+M/j14shefwvoWqTGcaazDLXM5JJLAn5m6k/Kp6mrn/BL/AP4JqfEL 9r7xjH/wUg/4KR3l34n1XW5lvfCPhnWhmEQ9Yp5IT8qRDgxQABcYYjpn5T/Yo8J+OP8AgtH/AMFT Ln4sfHiJ7rw9p1wdY1bTSSYbawhfFrp69thO1W4+fDk8tmv6A7KxttPtIrKyt0ihhjWOKONcKigY CgDoAK87ijM8XlEFhnK+KqRTqSX2IvanD+WK8rd3q9Pb4YynCY9+1jG2Gpv3IvXnkt6k/wCaV+9+ 2yFtbKCygS1tYFiijUKkcagKoHQADpU1FFfmdj9ISsrBRRRTGFFFFABRRRQAUEZBFFFAHCfH39nr 4Z/tHfDy7+HXxP8AD8d5ZzDfbTgfvrOYAhZom6o4z1HUZByCRXwtY+KvEv7KPiy7/Ys/bP04+MPh fq67NI1G9jLyWcJOI5Y2PzAKcEgHcjDKntX6SEZGK8Q/bx/Zw0z4/wDwPv0gslfXNDja90WYD5iy j54fo6gjH94Ke1fS8PZrDD1Vg8VrQm/nTl0nB7xae9t1vex8rxFk0sRB43C6VoL5VI9YTW0k1te9 um58R69oXxU/4Jo/E+HVvAWpjxV8KfFw3x6feES2moWrjmNgQVWYIeHA+YdQRkVl/tDfsqeGbXw7 Z/ts/sQapd2egmYXGoabp8rJc6Bcg5O3adyoD2/hzx8vT1P9j/UtO+Ovwu1j9i74szCWCe1kuPCN 7OMvZXCDcFUn+6fmA9N46Vy/7IHjjX/2YPjfe/Cr4hwhvD+q3raX4i0+5UNCj7iglweMc4J7q1fo 6xWNo1Ks1riaSXOvs16feS25rXV7Xvo7qR+byy/BYinTg7rD1G+R/aoVNNIvflvZ2va21nE6v9nT 4s+Cv259Lh8FfFWC10f4padBnT9ftoxCusqg/jC4HmDHOPqMDivd/g/+0/efDnxAfg78dNVkFzaS iCPUbsYkT+6JGP3wRjD+nXPWvkL9qr9nDVv2W/j4NU+HlzPZWE041Hw1dQud1uN2fLDHrsbjnOVx nOTXsX7S1v8A8NO/slWP7UnhW0WLxR4btzD4kt7Yf61I8CTjrgZEi/7LEV4+Y4HLsWqMov8A2atp HvSm+ivtF9uj0Pay7HZlg1WjP/eaOsu1SC3v3kuj6rU+6YbiGeJZ4ZA6OMoynIIPcVIDkZr4W/4J Z/t2j4g3R/Zz+IWrebewQtJ4au5nG6SIctbknklRyvfGR2FfdCkFQR6V+fZxlOLyXHywuIWq2fdP Zn6Hk2b4XO8BHFYd6PRrqmt0fzp/8HBPxxPxe/4KK+IPDlpdeZY+CtOtdFhAfKiUJ50pHvukCn3S vNP2BP8AgqJ8eP8AgnNbeIofgh4D8FanJ4mkgbULrxVpl3PIixBgqRmC6hCqd5JBByQOmK/b/wCL X7Ev7M3in4kaz4n1v/glrpfiu81HUHubvxHLNowbUJXO55SJrlZMlifvAHiuVl/4J6fslTSNK3/B IHTwWOSF1TRAB9AL3iv03B8TZK8jpZfXw/PTjGKac6Vm1re3PffXU/N8VkWaLOKmMpVnGbk3dU6r tfpdQa2Pz7sv+DnH/goRq15Dplh8IfhG89xKscMa+HtUJd2OAo/4mXUk1+4nwu1XxhrXw10HWvH8 dkmuXekW82qpp0TxwLO0as4jV2ZlUEnALE+5r5G0f9gb9lrQNWtdd0f/AIJG2cF3ZXKT2s6atom6 ORGDKw/03qCAa+hD8YfjWYDan9kDxH5ZXaVHifRcYxjH/H5XyPESynHqmstw8KSV+b36d322k9j6 jIauPwPO8fWnVva37qrZd/sLc/nc/wCCtfxwP7QP/BQT4keOYbrz7S31xtM05w2Qbe1UQrj0+5X7 H/8ABvJ8CR8Jf+Cd+keKru1MV7411e51mdiP9ZESIoT/AN+4xV7Uv2A/2WdY1K41jVP+CR1nPc3U zTTzSavohLyMSzMf9N6kkmvdPAfjL4jfDHwfp/gHwH+xLrel6PpVstvp2n2niPREigiXoqgXnAFe xn+dwzHI6OXYWKio2vepTs1FWW0u+vyPMyPLngM5qZhiXKTle1qVbRyfnDtp8z4O/wCDi3xX+1v8 HPFngj4sfBn4neK9B8JzaVLp+pP4f1We3hjvfNLqZRGwGWQgAt1wQOlfHX7En/Bbf9qn9kjxHql7 48u734naZqsCK+neKfEdx5ltIucPDKwk2ZzyNpBx2r9s/Hnirx58UPDF54K+If7Deq61pGoRGK90 3Ute0OaGZCMEMrXZBr5Z8Q/8Ejv2NfEesPrVx/wTA8UWzSuWkt9P+Jtpbwn2EcephVHsoFTlOY5Z TypYHMKMZJdYzp3avfV88WvkzhzrLc1q5u8fllecb68sqdaydktF7OSa9Ufk/wDtvfthfET/AIKO ftA2vxFl+E2maRfSW0enaVonhu2aaWUbiRvfG6eQluuAMdAOc/Q37Qn7Ofin/gnJ/wAEjH8M/EaL 7D4++OPim0bUtNLDzLDTbZGlW3cdmOSzjoCyr1U1+mv7P/7Mvwq/ZfuV1H4Jf8EyJtHvlGBqZ1/R Z7rH/XaW9Z/1rW/aK+Efhv8AavTTR+0L/wAE6dR8UjRjIdMXUfEujN5BfG7bi9HXA/KuuXEdCNbD 0aVNRw9Npte0p8z5dUvita9nu35nNheHJqNfE4mpKeJqJpP2VVRXNo2/cWttNkkfiT/wRG+BbfHP /go14C068s/NsdBuZNbvQy5QrbIXAb2LbR9cV+2X/BSGa81ez8M+DIGYwGSe8uEHQsoVE/8AQpK+ WPh1+3X/AME1f2Evi3rtv8MP2FfEHhHxZYyT6NrMmnxWbOPLlHmRbvtJBG9ByDzivoD9pT406P8A E/wj8PvjLo+my2+neKvDX22xhu1XzI1Yq+1tpI3AOM4JFdeb/wBq43iTDY2th5U6XK1Dm5Xd8rk2 rN9NV5CyjE5LSyHE4LDYmNSqpJzUbqy5lHW6XXR+Z0n7FPhO3+GPwN8Y/EkW4F3OJFV8YIjhiJUA /wC87fpXg/wc+FbeNfjVoNprURlW91uOW93D76hi8mc+oB/Ovoz4H+I4Nf8A2QPFJ08Kz2sV6kiL /e8sMB+Rryj9mvx7p7/HTw3CyoPNvjEpPq6Mo/nXm4aviVUzGql792vko6Hp4ilR5MvpP4bJ/Ny1 Nj/goFLeeMfivaeHlBNrpGnqI4wfl3yfMx+vQfhXgf8AwXZ+IMv7Ov8AwTJ8Efs/eH5DbzeNNSgG o+WcGSKJPtEqt6gyMn/fAr3j9rXxRFofx41G1vIV4SFxuHVSorhv+Cxn/BN79on/AIKLaF8NpPgT qmgQ2Wg6fLJdrrOoPDlpUTaU2owPyj2qcDWw2FllixMlGlH323tflur/ADLxdLEYl5i8NFyqv3Ul vbms/wADyH/g190jwH4S+EnxI+JWv67ptnqOoa3a2MRvLuONxDHGzsBuIOCzL+VfqkPib8PMc+O9 F/8ABrF/8VX4XWH/AAbcf8FG9LiMOl+OfB9sjHLLb+JJ0BPuBEM1Mf8Ag3I/4KWk5/4WX4X/APCq uP8A43Wmc5Zw5nGaVMXLM4x52tLN2skrX+Q8ozDiHKsup4WOXyfL1va+t/1P3L/4Wb8PP+h70X/w axf/ABVH/Czfh5/0Pei/+DWL/wCKr8NP+Icj/gpb/wBFL8L/APhVXH/xqj/iHI/4KW/9FL8L/wDh VXH/AMarzP8AVnhr/oax/wDAX/mel/rDxH/0Lpf+Bf8AAP3L/wCFm/Dz/oe9F/8ABrF/8VR/ws34 ef8AQ96L/wCDWL/4qvw0/wCIcj/gpb/0Uvwv/wCFVcf/ABqj/iHI/wCClv8A0Uvwv/4VVx/8ao/1 Z4a/6Gsf/AX/AJh/rDxH/wBC6X/gX/AP3L/4Wb8PP+h70X/waxf/ABVH/Czfh5/0Pei/+DWL/wCK r8NP+Icj/gpb/wBFL8L/APhVXH/xqj/iHI/4KW/9FL8L/wDhVXH/AMao/wBWeGv+hrH/AMBf+Yf6 w8R/9C6X/gX/AAD9y/8AhZvw8/6HvRf/AAaxf/FUf8LN+Hn/AEPei/8Ag1i/+Kr8NP8AiHI/4KW/ 9FL8L/8AhVXH/wAao/4hyP8Agpb/ANFL8L/+FVcf/GqP9WeGv+hrH/wF/wCYf6w8R/8AQul/4F/w D9y/+Fm/Dz/oe9F/8GsX/wAVR/ws34ef9D3ov/g1i/8Aiq/DT/iHI/4KW/8ARS/C/wD4VVx/8ao/ 4hyP+Clv/RS/C/8A4VVx/wDGqP8AVnhr/oax/wDAX/mH+sPEf/Qul/4F/wAA/cv/AIWb8PP+h70X /wAGsX/xVPt/HfgfWJhpdj4u0q5lnyqwQ6hG7PxyAoOTxmvwW+JP/BAT/gon8Lfh9rfxJ8R/Erw2 dP0HS57+9EPim4L+VEhdsAxjJwDXnf8AwQ3tvGHjr/gqJ8LrBvEGoT21ndahfXYku3ZBHDp1y43A kjBfYv1YVu+C8tq5dXxeFxyqKlFt2j2Tdr362MP9bsxpY+jhcTg3B1JJK8u7Svt0P058ReD7r4M/ tK3Gt6BEYzo3iP7TAiDA8svv8v6FW2/Suo/bs+E9nH8ZV8X6ZAoj1ywjuHKj70ijaW/ILR+0/wCN tPsfjtr1mscZMV0iuSO+xc/zrt/20tft9E0/wat4i/aH0cbww5Hyrn9a9CnicT9cwVb7UoNPzXKp fn+Z5lShh/qmMpfZjNNerk1+Qvx38Mj40/se+GfHGoRCTU9HjiWWZ/vNtzDIT9Sob8aw/wBg2wiF x4o+FOvQmTTNd0pvNgboxwUYD0yjH8h6V3Hg7XEuP2GNT1m5RBEEmaMEcYEygfrmuB/Yw8WR6t8c bOxt0Ub7G4L7R2CZrzo+0eTYzDr4YTk15WtLT52PQn7P+18JW+1OMU/O6a/I/LCT4geKv2Z/2l57 3Q7x4b7wf4tkWMx8Z8icgr+Kgj8a/oE+GvjGw+Ifw90Tx3pkitb6vpcF3EVPGJEDf1ry/wARf8E7 v2LPGXiC98W+J/2ePDt5qOo3T3N7dzWYLTSuxZnJ7kkmvWfCHhHw54C8L2HgvwhpMVhpel2qW9hZ wLhIYlGFRR2AFZcZcUZZxNQw8qNKUasFaTdrNNLtrutL9zHgThHOOFMRiY4itGdKo04pXumm+6tq nrbsaJRCNpXik8qL+4Pyp1FfBrQ/RrIb5UX9wflR5UX9wflTqKAshvlRf3B+VHlRf3B+VOooCyG+ VF/cH5UeVF/cH5U6igLIb5UX9wflSPFHtOFAOODin0EAjBo3CyPx5/4K5/so3PgX9q+/+Iem6d/x KvGsK6jDIifKt0FCXCZ7ncBJ/wBtR6V6z4LvLj4lf8EzvDmqWAL6p8MNSk07UYl+8tox+Vsemwwn P+w3pX23+1b+zb4f/aR+GU3hO+ijjv7VjcaRdsv+qmA6H/ZYcH618gfs2aFqv7O3j/WPh98Q9CkP h/xBCdO8S2Ei8AchZQO5Xcee4Y1+s5fxA834eo02/wB9hXF26yik1+MW1627o/H8VwxDJuJMRXir UcUmm+ilJqX4SV/Q3f8Agmd8YNK13W/EvwH1+7VU1/TjcaernhpVUpIo9SUYNj0jNeFa54j8QfAL 44SaXfK8d54Z15WQSH73lyBlb6EAH8a534++CfiP+wl+0PZazod47WUV2uo+E9ZjJ8u7t92QuehI Hysv9DXsX7XOg6D+2J8EbL9tT4D24l1HTrVYPHehW3M8BUcy7Rydnf1QhuitXrRo4ahmccUmnhsW kubop7K/bmWjv1VmeZLF4itl0sI9MThG3y9ZQ3du/K9dPs6o7T/gpDDDenwr8f8AwvIZdJ8RaWkb TpyFcDegJHQ7Wx9VPpX0/wDsYfEyw+K37OfhzW4bgNPaWYsbxQ2SskICc+5UKfxr4t/YV+MXgv8A ak+CmrfsM/FrVxb37wPN4N1CUjKuMsFTPV0bnb/EpYetW/2DPjX4l/Y9/aN1X9lP4/IdNg1a7Edn cSt+5W6ziORWPBjlGAD64zz0+czPJ69TKamXtfvsK+aK6ypPqu9lvba2p9Flee0Kea0swUv3GJXL J9IVF0fa72vvc/RxMbeM9e9LSIdyg0tfmVkfqS2CiiiiyAKKKKLIAoooosgCiiiiyAKKKZPNFbwt NNIqIqks7HAAA5JPaiyA+V/+C0vx50/4A/8ABOX4i6vd3ix3niHSv+Ee0mPdhpLi8zEdvusZkk+k Zr8+P+DZP9ntrbxt4/8A2y/FkAh0rw9ojaLpd1MMIZJNs1y4J4GyOONc/wDTRh2rB/4K8ftOeNf+ Cp/7aPhn9hn9lON9a0bw7qz20Elscw32pMNlxdsw4EMEe5Qx4A8xs/NX1F+0V4i+G/8AwTq/Y80D /gnx8F9RSfW57BT4s1C3IDEP80zNjkPM+QB2jGD1FfqmXZdicNw9DKqa/f4tqUl/JTX2n2079W1u fl+Z5nhquczzSo/3GFXLF/zVH0XfV9OxheFdf1L9p/8Aapi03T1eSPxD4kMsnrHaB9zMfTESn8RX cf8ABSf4xWmpfHyPwXpNyHTQNNjgkVD92V/mK/UDb+dTfsm+GtG/Yg/Zz1f9sj432wh1rVrEw+FN GnOJnVxlFweVMjAEn+FFzj181/Yf+GutftNfGjU/2kfjLexw+FdAvn1bXtUvm2w3FwD5giyeNq9T 2AUDvXq82E+vVMd/y4wseSL/AJpuyaj30Sjpu7nhOpiXgqeC/wCX+JlzyX8kFs5dtW3rsrH0Z+0n 4iH7PX7CHhf4ZalceTrGtQQi4tyeVLEzzA/7rNt/Cue/4JWaHe+LPiBrvj+eMm00uwW2jk7GaU5x /wB8qfzFfNH7Wf7TviP9tD9o1NL+Gul3V1YJcDTfCmnKDvmXdjzSOxdvm56LjPQ1+mP7Hv7Pll+z d8DdK8AO8c2psputcu0HE13IAXwe6qMIvso7k14ecwlkXDPsa+mIxLcmuqUrN/hZep7mSVln3E3t cPrh8OlFPo3FNL9Weor0paAAOBRX5mfqIUUUUAFFFFABRRRQAUUUUAFFFFAARkYrzH9of9nfT/jH ozXui3qadrtvHi0vcfJJ6JIB1Hv1FenUjdOlbYfEVsJWVWk7SRhicNRxdF0qqvFn5yfE3xHBpGmT /sv/ALcHgq8j0SSUvo+vQx75tLl6LcW8nSSP1UduMdq8D06b9on/AIJnfEy3+KXgu9g8V/D/AFht i6rYsZdN1m1P/LKYD/UzAHo2CD0yK/XP4ofCb4e/GTw3J4R+JHhW11SxkziO4jyYzj7yN1Q+4r5D 8ff8E5/jl8FW1DUv2TPHNprugX4P9qfDjxmBLaXsfdFZhtLehbaQcHeMV+l5HxTl1SnLD4hRiqnx wlf2cvOLSbpy9bx6uz1Py3iLhPMqdSOIwrlKVP4Kkf4kbdJRbXtIryakvNaHzZ8RfgZ4N/aOs/8A hrT/AIJ26zJDrGnyi+1/4exTiPUdKuQd5ltlGN65BO1ev8OT8teg+Cvi78EP+CoHgmD4K/Hy7j8F fGbRoTFo+uSQ+UL11HTBxzkfNEcHqV9K8h8U/BjSfhz8SofE3hG78Q/s3/EaGUtb2HiVZToF7Jnk QX0YZYoye0m6PGPmNb3xR1fw38Qo7Qf8FCfgTq3gvxA4U6V8bfh5AtzZ3TDlJZTbFklHRtyEuP7o NfVVaKr04ck5S5NaU1Z1oeVr8uIp/wCFt23Tep8HRxdXD1qvtIRXPpVg7+xn/e/mw9R/34xjfZpa H1T+zP8Atd/Ez9n7xLZ/sqft5WjaTqsWLfwt44lctYaxEPlRWmxgPjABbBPRgD1+x7eaKeBJ4ZFZ HQMrK2QQRwQe9fnZ4M8f/GbTvhiPB/x28J6R+0t8H5I9kHjHwXJHeanYRAYDTW2fN3KOuOVxy3av T/2dtd8RaBpA1P8AYn+OGnfEnwhAf3vw58W6i1vqulLz+6hmlAkiI7R3C49GAxX51neTU6tSVeHL CV9bfwm352vRk/5Kiit7NbH6fw9xDVoxjh6nNUjbRS/jRXmrtVorpUptu26b1f2PRXlvgX9qX4fe KdUi8I+LIb/wd4lcYfw54ttTaTsR1MLkmK6X/ahdx646V6etxA6CRJkKnowYYr4+th62Hly1ItP+ tu/yPvcNjMLjIc9GakvJ7eTW6fk7ND6KQOhxhxz05pcg9DWJ0hRRSFlXqwH1NAbC0ZHrTZJookMk sqqqjLMzYAFfNH7Uf/BVD9l79nHWn+H2j6ve/EDx47+XZ+A/AVsdRvmmxwkvlZWDnHDndjoprow2 ExWNqezw8HKXZL8+y82c+JxeGwdPnrTUV5s+kdU1XTNE06fV9Y1CG1tbaNpLi5uJQiRqBkszHgAD ua/Kj/goZ/wVF+K/7afjmT/gn/8A8EuNIv8AxFd6mzWvirxtpwKQiPOHjglPEcIGfMuGwDyEyPmZ v7RWl/tmftoxR63/AMFAPi7Z/s//AApnbdY/C7Qrv7X4h1tc8JJDDudmOQNpGQc5jXrXoXwn+Bvx M8KfC6b4bfsffCay/Z4+Gbxbtc+I3jt0j8QavGBzJ5ZbfEMcjzCoGSQeor7bKMmweWtYnEuM6i2u /wB1F921rUkukKaeu7PhM64grYq+GoRlFdVFfvZL00VOL6zqNKx5j8JfCHwJ/wCCMvw5m8BfDu5t PiD+0Z4rtBbarqFlF50ejh8HyUwMgBsEIPmkIDNgAA9T8H/2YtB+BtnN+3L/AMFMPFHlXk05vNK8 J3rCW7vbk/Mhkjz875xtiHCjl8AHG98KrT4MfAg3h/YZ+EV/8V/G+X/tj4u+LiLfRtPk6vJ9rnKo 5BycRckdGbkHyzxTrngnxn8Tx4l+JOt6z+018UyxFj4T8HWsx8O6W2fumVV/fKp7RjacckGvrqKq 1/aRg5pz1qTatWqL+WKfu0KfZzadtUmz4HFYtRcKlXkly/wqabdCD7uS1xFTvGmmk92kaHiS5+L/ APwUm8byfHP4zazH8O/gv4bc+XqGpyiKKGAHlIAf9dOwABYDAyBzgKea+M37Tuv/ALSiaX+xZ+wd 8OdTh8E2biG3tbWErc6w4PNxcE42IT83zkerY6D2jR/+Cbn7an7Z+rWGvftm+ObbwR4Us9v9l+BP DoQ/ZYh0RY4yYojjuS7euDX27+zl+yn8D/2XPDI8NfB/wPb6cHQC6vnXfc3RHeSQ/Mfp09q4cfxH kmUKCgo1p0/4dODbpU3/ADSn/wAvJ+a0WtrPU9DLOF+Ic7nN1OajTqfxKtRWrVV/LGH/AC7h/der +0pLQ8Y/4J6/8E4NC/ZV0mPx78RZrbVfHN3APNniG6DTVYcxQkj5m7F+M9uOv1dEML0xz0p1Ffl+ Y5ljM1xcsTipc05fguyXRLoj9gyvKsDk2DjhcLHlhH72+7fVhRRRXCeiFFFFABRRRQAUUUUAFFFF ABRRRQAUUUUAFFFFAGZ4o8JeG/GWlSaD4s8O2Wq2My4ms9Rtkmif6o4INeP3f7BHwc0mS4uPhRNq /go3W77RZ6Dfn7DPnr5lpLvhYHuNooorpw2NxmDuqNRxT3V9H6rZ/NHBjMtwGNadempNbNrVej3X yZxX/DubStF8SHxT4WurPS9T3ZbWfCbSaJdyHqN6wM1uwHXBiwa3tQ/Yt17V7231vxPreja7qUKg prV3pxstWiPoL2yaNpD7upHtRRXfLiHNXJN1NUrX627LsvJWR5sOF8lgrRp2V726X726vzdzs7H4 IeMrnw4fDvijXrXX9NY/8gjxbax6go9/NCozEdsjj1qxonwEu9ABbwb4l1Tw2V6WtrqjXlox9fJu Ado9FDAD0ooriePxLTV9PRW+7Z+rVzvWWYRNOzuut3f7916JpG/Y2vxl0VwLu60TWYU4UmN7WZvc kFkH0C11OnS381ssmo2SQSkfNGk28D8cDP5UUVzOfPrZL0Vv+AddOn7J2Um/V3/4JP8AhXE+NZ/j zfSvY+ANL8OWA6JqOsXU1wCPXyYwhB+rkUUURai72T9S6kedWu16HkXjH9ij4pfGppIv2gf2jvEu s2Mp+fQNBvBo+nEf3XW2AlkUjghpOe9bXwz/AGHfh58FtDOh/CbQtO8J2jrtmg8J2MVpPKO4e7cP O4PcMx68UUV2vM8a6apqVo9klb7tr+e5wPK8G580k2+7bv8Afvby2NbQf2Y7bwjqE+t+A9I0XRtS uBmXXZbU3+puem1rm5LEg89MAdhXP+Kv2M9T8fagmpeMNdt9VnRt0Vx4mL6mbdgesduxS2H/AAKM ketFFOGbY6lLnhOz79fv3fzuY1MhyyrD2coe7vbpfulsn5pJli5/YB+Fvi2SEfGfxFr/AI1gt8CD SNUv/I0yNR0C2lvsjGO2QcV614B+GHgD4X6Wmg/DrwRpmh2SHm30uyjhVuOp2gbj7nJoorLE5hjc VDkq1G43vbaK/wC3VZfgdOFyrLsFN1KNJKT3la8n6yd2/vOhooorkPQCiiigAooooA//2VBLAwQU AAYACAAAACEAlrWt4pYGAABQGwAAFQAAAHdvcmQvdGhlbWUvdGhlbWUxLnhtbOxZT2/bNhS/D9h3 IHRvYyd2Ggd1itixmy1NG8Ruhx5piZbYUKJA0kl9G9rjgAHDumGHFdhth2FbgRbYpfs02TpsHdCv sEdSksVYXpI22IqtPiQS+eP7/x4fqavX7scMHRIhKU/aXv1yzUMk8XlAk7Dt3R72L615SCqcBJjx hLS9KZHetY3337uK11VEYoJgfSLXcduLlErXl5akD8NYXuYpSWBuzEWMFbyKcCkQ+AjoxmxpuVZb XYoxTTyU4BjI3hqPqU/QUJP0NnLiPQaviZJ6wGdioEkTZ4XBBgd1jZBT2WUCHWLW9oBPwI+G5L7y EMNSwUTbq5mft7RxdQmvZ4uYWrC2tK5vftm6bEFwsGx4inBUMK33G60rWwV9A2BqHtfr9bq9ekHP ALDvg6ZWljLNRn+t3slplkD2cZ52t9asNVx8if7KnMytTqfTbGWyWKIGZB8bc/i12mpjc9nBG5DF N+fwjc5mt7vq4A3I4lfn8P0rrdWGizegiNHkYA6tHdrvZ9QLyJiz7Ur4GsDXahl8hoJoKKJLsxjz RC2KtRjf46IPAA1kWNEEqWlKxtiHKO7ieCQo1gzwOsGlGTvky7khzQtJX9BUtb0PUwwZMaP36vn3 r54/RccPnh0/+On44cPjBz9aQs6qbZyE5VUvv/3sz8cfoz+efvPy0RfVeFnG//rDJ7/8/Hk1ENJn Js6LL5/89uzJi68+/f27RxXwTYFHZfiQxkSim+QI7fMYFDNWcSUnI3G+FcMI0/KKzSSUOMGaSwX9 nooc9M0pZpl3HDk6xLXgHQHlowp4fXLPEXgQiYmiFZx3otgB7nLOOlxUWmFH8yqZeThJwmrmYlLG 7WN8WMW7ixPHv71JCnUzD0tH8W5EHDH3GE4UDklCFNJz/ICQCu3uUurYdZf6gks+VuguRR1MK00y pCMnmmaLtmkMfplW6Qz+dmyzewd1OKvSeoscukjICswqhB8S5pjxOp4oHFeRHOKYlQ1+A6uoSsjB VPhlXE8q8HRIGEe9gEhZteaWAH1LTt/BULEq3b7LprGLFIoeVNG8gTkvI7f4QTfCcVqFHdAkKmM/ kAcQohjtcVUF3+Vuhuh38ANOFrr7DiWOu0+vBrdp6Ig0CxA9MxHal1CqnQoc0+TvyjGjUI9tDFxc OYYC+OLrxxWR9bYW4k3Yk6oyYftE+V2EO1l0u1wE9O2vuVt4kuwRCPP5jeddyX1Xcr3/fMldlM9n LbSz2gplV/cNtik2LXK8sEMeU8YGasrIDWmaZAn7RNCHQb3OnA5JcWJKI3jM6rqDCwU2a5Dg6iOq okGEU2iw654mEsqMdChRyiUc7MxwJW2NhyZd2WNhUx8YbD2QWO3ywA6v6OH8XFCQMbtNaA6fOaMV TeCszFauZERB7ddhVtdCnZlb3YhmSp3DrVAZfDivGgwW1oQGBEHbAlZehfO5Zg0HE8xIoO1u997c LcYLF+kiGeGAZD7Ses/7qG6clMeKuQmA2KnwkT7knWK1EreWJvsG3M7ipDK7xgJ2uffexEt5BM+8 pPP2RDqypJycLEFHba/VXG56yMdp2xvDmRYe4xS8LnXPh1kIF0O+EjbsT01mk+Uzb7ZyxdwkqMM1 hbX7nMJOHUiFVFtYRjY0zFQWAizRnKz8y00w60UpYCP9NaRYWYNg+NekADu6riXjMfFV2dmlEW07 +5qVUj5RRAyi4AiN2ETsY3C/DlXQJ6ASriZMRdAvcI+mrW2m3OKcJV359srg7DhmaYSzcqtTNM9k Czd5XMhg3krigW6Vshvlzq+KSfkLUqUcxv8zVfR+AjcFK4H2gA/XuAIjna9tjwsVcahCaUT9voDG wdQOiBa4i4VpCCq4TDb/BTnU/23OWRomreHAp/ZpiASF/UhFgpA9KEsm+k4hVs/2LkuSZYRMRJXE lakVe0QOCRvqGriq93YPRRDqpppkZcDgTsaf+55l0CjUTU4535waUuy9Ngf+6c7HJjMo5dZh09Dk 9i9ErNhV7XqzPN97y4roiVmb1cizApiVtoJWlvavKcI5t1pbseY0Xm7mwoEX5zWGwaIhSuG+B+k/ sP9R4TP7ZUJvqEO+D7UVwYcGTQzCBqL6km08kC6QdnAEjZMdtMGkSVnTZq2Ttlq+WV9wp1vwPWFs LdlZ/H1OYxfNmcvOycWLNHZmYcfWdmyhqcGzJ1MUhsb5QcY4xnzSKn914qN74OgtuN+fMCVNMME3 JYGh9RyYPIDktxzN0o2/AAAA//8DAFBLAwQKAAAAAAAAACEA87ZKt0T2AABE9gAAFQAAAHdvcmQv bWVkaWEvaW1hZ2UyLnBuZ4lQTkcNChoKAAAADUlIRFIAAAKhAAABxAgGAAAACXPzKgAAAARnQU1B AADZA0LWT6EAAAP4aUNDUEdlbmVyaWMgUkdCIFByb2ZpbGUAAHicjVXdb9tUFD+Jb1ykFj+gsY4O FYuvVVNbuRsarcYGSZOl6UIauc3YKqTJdW4aU9c2ttNtVZ/2Am8M+AOAsgcekHhCGgzE9rLtAbRJ U0EV1SSkPXTaQGiT9oKqcK6vU7tdxriRr38553c+79E1QMdXmuOYSRlg3vJdNZ+Rj5+YljtWIQnP QSf0QKeme066XC4CLsaFR9bDXyHB3jcH2uv/c3VWqacDJJ5CbFc9fR7xaYCUqTuuDyDeRvnwKd9B 3PE84h0uJohYYXiW4yzDMxwfDzhT6ihilouk17Uq4iXE/TMx+WwM8xyCtSNPLeoausx6UXbtmmHS WLpPUP/PNW82WvF68eny5iaP4ruP1V53x9QQf65ruUnELyO+5vgZJn8V8b3GXCWNeC9A8pmae6TC +ck3FutT7yDeibhq+IWpUL5ozZQmuG1yec4+qoaca7o3ij2DFxHfqtNCkecjQJVmc6xfiHvrjbHQ vzDuLUzmWn4W66Ml7kdw39PGy4h7EH/o2uoEz1lYpmZe5f6FK45fDnMQ1i2zVOQ+iUS9oMZA7ten xrgtOeDjIXJbMl0zjhRC/pJjBrOIuZHzbkOthJwbmpvLcz/kPrUqoc/UrqqWZb0dRHwYjiU0oGDD DO46WLABMqiQhwy+HXBRUwMDTJRQ1FKUGImnYQ5l7XnlgMNxxJgNrNeZNUZpz+ER7oQcm3QThezH 5yApkkNkmIyATN4kb5HDJIvSEXJw07Yci89i3dn08z400CvjHYPMuZ5GXxTvrHvS0K9/9PcWa/uR nGkrn3gHwMMOtJgD8fqvLv2wK/KxQi68e7Pr6hJMPKm/qdup9dQK7quptYiR+j21hr9VSGNuZpDR PD5GkIcXyyBew2V8fNBw/wN5doy3JWLNOtcTaVgn6AelhyU42x9Jld+UP5UV5QvlvHJ3W5fbdkn4 VPhW+FH4Tvhe+Blk4ZJwWfhJuCJ8I1yMndXj52Pz7IN6W9UyTbteUzCljLRbeknKSi9Ir0jFyJ/U LQ1JY9Ie1OzePLd4vHgtBpzAvdXV9rE4r4JaA04FFXhBhy04s23+Q2vSS4ZIYdvUDrNZbjHEnJgV 0yCLe8URcUgcZ7iVn7gHdSO457ZMnf6YCmiMFa9zIJg6NqvMeiHQeUB9etpnF+2o7Zxxjdm6L+9T lNflNH6qqFyw9MF+WTNNOVB5sks96i7Q6iCw7yC/oh+owfctsfN6JPPfBjj0F95ZNyLZdAPgaw+g +7VI1od34rOfAVw4oDfchfDOTyR+AfBq+/fxf10ZvJtuNZsP8L7q+ARg4+Nm85/lZnPjS/S/BnDJ /BdZAHF4LriPvQAAAAlwSFlzAAALEwAACxMBAJqcGAAAACR0RVh0U29mdHdhcmUAUXVpY2tUaW1l IDcuNy4xIChNYWMgT1MgWCkAZAOCGAAAAAd0SU1FB90GEQ0mN3Yq1XMAACAASURBVHic7J0JvFTj G8efiEgphJSQkqUiRdnXJKSFSoVEKfuWKFuWSqgsERFlC0W2shXZEhGhkqUiSvztZM/8z/d97zv3 3Lmz3jszZ5bn+/mczyxn5sw7M2f5vc9aJeQhiqIoiqIoipJF1gl6AIqiKIqiKErxoSJUURRFURRF yToqQhVFURRFUZSsoyJUURRFURRFyToqQhVFURRFUZSsoyJUURRFURRFyToqQhVFyRlWrlwp//77 b9DDyHl+/vln+f3334MehqIoSqVQEaooRcpnn30m5557rnz11Vfh55544gl54403AhtT9+7d5YMP Poi5/u+//5YbbrhBjjvuOBkyZIj88MMPlf7Mhx56SG677bZKbYOxnHnmmTJo0CB58cUXKz0mx2mn nSaLFi0q9/xVV10l9957b9o+R1EUJQhUhCpKkfL+++/LpEmT5Lzzzgs/9/rrr8uHH34Y930Irttv vz3Tw4vKsGHDZMGCBXLppZdKkyZNZPny5ZXe5h577CEHHHBApbZx/fXXy7HHHitt2rSRfv36yaOP PlrpcUG3bt1k6623Tsu2FEVRco2qQQ9AUZRSfv3VLj/9JPLbb3b5808RPK9r1oj88Yd9/M8/duHx X3+JrF0rQu+z//6zC6+56y6R//1PpE6d2J/XoUMHefPNN431c5999imz7k9vIzfeeKN89NFHst9+ +8mpp54qr7zyirzwwgtSo0YNIwBPOOEEmTt3rgwYMMC8Z/DgwXLhhRd6n1nHbBP3OkKK1956663y /fffS7t27aRXr17m9WPGjJFWrVrJnXfeKZdcckn4s3HJX3311Wb7iE3H4sWLpW3btrLrrruaxfHp p5/K2LFjzZhPPvlk2Xvvvb3f8VfvN7hLatasKW+99ZaceOKJZjzus2fOnGlev/HGG8svv/wizZs3 N+8ZPXq0sRIfddRR0rNnz6i/Q5UqVcr9lgceeKCsu+665rvOmDFDunbtWu77NW3a1Kx7/PHHZYMN NpD+/fub78H2jzzySNlxxx3NtkaOHCm9e/c2E4WddtpJatWqJe+8846MHz/eiNI/+OPj/E+ML9pn x+PMM+1+06CBSO3aIuutJ7L++naJdj/V9YqiKJGoCFWUNIOH+OuvRVatEvnmG/GEl8i339rH3GdB XH73nV2PgIyGp52MGPA0krmQV6tml+rV7eJpGFlnHe8grlq68LqPPxZPKNptxBOgUNV70/Dhw41w jHTDI+ZatmwpF198sQwcONAb51o55phjjCjacsstjet88803l6FDhxrhs2LFCrnuuutkl112MQIK wYSw+slT1AcddJARRdttt52cccYZRjidcsop8thjjxmr4RVXXOF919rmc+kkzPZ47Beg0LdvXyMi v/N+vHPOOccISLaPYLznnns8Qf6XGRfCeB3vx2G7CFkEIWM96aSTjLBEpCH0sKhiWV29erUcffTR cvzxx0u9evXk/PPPlylTppjXRvsdTj/99Ji/6cKFC8NiMvL7TZs2Ta699lpjSf7yyy+NIOfzEcEI 5lGjRnn7yrdGUPOf8H5+OwQrEwZeQwgCr8P6Gut/YnzRftt4MMkhkoB9dfvtRZiTsH/9/bed8HDr vx/tuWjr2b/Zzqab2mWTTUrvR1v869n3FUUpXFSEKkoSYF384gu7fP65XVautIunvcyCBdOBIPR0 mmy1lQjeVC6ongbyBJy93WILKzC54LpbhGU6cEY6LKPJQHwlljS/C5nYy6eeesoIPix7u+22m7GA Im7q168v22yzjXFjA8Ly7bffNrGcCLlnnnnGWB1feuklI6aefPJJY6HDXQ2487HOIUIBUYYV0YFo wtrHmCI54ogj5LXXXpMrr7zSCNRnn31WlixZIg0aNDBiFBgbr2GbiFTiPbFQwqbeH4Hld3tPZWHt ROC58AOEMdvDWsv7sCLG+x0i4buv8mYafD6xtg7/9yP84bLLLjO/HQu/+XPPPWfE7yGHHGLiXRGP iGYmCA6stgcffLD5DHj33XeNWE80vsjfNh7e0AyIRiI0br1V5NVXRfbfP6m3xwVLPpMzlh9/LL3v Fo6naOt5XzyRWreuePujPcZYEk26FEXJLVSEKoqQbWwtiJ4ukU8/tQsXRkIOsWA6uNBx0cNlue22 IniE3XP16lnRWaJ38gasgrigEYVYLgHLHLikJYRm+/bto76/Y8eOxsX8sfcD4kLv1KmTcX/j/kXM /eipis022yz8elzLfncy7nI/O+ywg8yZM8dYNath+o2gWbNmRrwhREeMGGHEG2IMyyJg+dx3333N /eqesl/X94ew7uGHHzaWSsQ31lKHG+fGPvNbKr8DYncrbwfAaunH//34DISw/7cgy71x48bGAos1 Gguss3I6yIb3WzP5TvxvicYX+dsmAz+XN3eQzp0JMbDitHfvlDdTBme9TzW8lUIJkYLVL1IXLBBP hPP97YJodYKUheM08jGTQEVRcgMVoUrRgNVy8WIRko0xfn3yiV1wP4KnBzwBZMUlt+3acVG3yzbb BDny5Bkzxt4mawV1YKlE3E2dOtVYIonpRCxhKdx5553Na9aWxA2sv/76Rkw5EJ2Iu//++09atGhh LIiXX365cd0D1rkJEyYYoch7sfxFxp/6wZWPMCNhKjIBat68edK6devwOBBmbB83NS58rIehki// NTEREfTo0cMkDyHWiFH1Q4gB35HYSr4zbnV+k1i/QyRYQNdNMAMh/vP555+X/fff31heX375ZTNu 4DfE+vvNN9/InnvuWeZ9hEAgcvlsxDnhBoj1eP9TZTn0UHu8eJsW5hBHHZWWzaYExmC8BizJgAh1 gpSFeQnH+rPPlj7GYxFPpHKLF0NRlMyjIlQpKIhB46KDheS990So9sOFlDhMQEziEmfxrvmeyBAh fK9Q3HgDByb/WqyAfksg1jcEpLPkIey6dOliXNdY4kg+QiAefvjh0rlzZ5OkNGvWLGPxXLNmTdj6 SOwiwmpSiX8XdzAiicQfLI2bbLKJTJ48OeoY3ONbbrnFCFUsnrimHTyPlRVrJYIT1zWue6yhiNGG DRsaEUepKbYTmUCEtZGYVWJAeT3wGhZejzAlTpNtYkXl+8X6HfzEEp+R348YVMQ5ZZzYFtvZfffd zToss8R0IuAj30/2PuNmzIhOvqf7brHGF/nZFYHj5OmnRTCQP/mkjRPNZbC2EkYcEUpcBpKv/CKV W2/eITNnlj4mPrZxY/v9OT/4bzfaKHvfR1EKnSqhUKo2E0UJHow9778v8vbbdkFsvvOOtQCioTwt JSQDozNYEJtYOgudVONBsZr5BRSng0jhRhIQbuP1fCnOWPGwfFYvCWRlO37R988//5R5PWAJ5fmN fFfxyM/3P2b7ECmksJKynVoRfyjPIcLq+GYUZNn7Yyvddvme7nO4z+I+hzHg5kYsJ/odYn2PRM+z fX6HyHWR4418/2+//WYqE7jTtv+/ihxfrM+uCCQs9e1rj7liOI6IFkGYLlliw3TcLQtW4WjiFCtq lKIJiqLEQUWokvNQZojEbZZ58+yC241McLyWGJLw0LZoYYVnJY0/eQ0XQeL5zjor6JEohUb//nb/ Gj8+6JEEB1dLkhAjxSm3RKggSCPFKbfpSjpUlEJDRaiSU+AOmzNH5JVXRObOtZYXoGQMgnOvvUT2 3tsKT0SoUsprr4lQc12PaCUTEEvJJO+ee0Tatg16NLkHZdeiiVOSHF1lDM5beGlatrTufkUpdlSE KoGB5QCxOXu2LQVDHCeQX0HsGQkRiE49WScHyVRk9+sRrWQKjtWTT6YWqkiNGkGPJj9w5d1IgiR0 aP58SmzZ8x9VzhCkCFMWjmF16SvFhIpQJWtwAia2jASAl16yz5EodPDBVnBixcPiqVSMVONBFaUi 9OhhQ18GDw56JPkNyZLEsSNIEaYsNLBwotTd4s4v5hAjpbBREapkBDJQZ80Sef55u+CSAtx4lH7h tqTWuZImEKEbbmjjZZXo0MWHmrCFUg0hCLDo4alYtkw7GqUbrKNOkDpxunq1deP7LaZ4i1SYKoWA ilAlLRAPRS2+6dNtSRdOplzoqS2I4KR+tl74MwsilBrngwYFPZLcheLn1IqnEYGvZrySIjS7op7m 1VcHPZLCh0mTX5SyUEYKazQTeTpa0VQg2VqqipJLqAhVKgQlTJ55xnYroX4gJ0ouSh060L7QWjuj NLtRMggilP8lomGPEgH7aJ8+Ir7yo0qKEOOIZY7kG51cZh+SxJwoffllG1NPgX1CmtxC0w1FyXVU hCpJwV7CyY724o8/Tjcae9JDcHbpYrsLVdXWB4Hx8MMiPXtqPGgy3HSTzVq+446gR5LfnH++rVBx 3XVBj0ThuKdJB2LULRgB/KKU2FJFyTVUhCoxoW864mbqVNt9iBik7t2tBYkOKsQfKrkB/wfhEHo0 J4bMbvqiU0lAqThLl9pyadTNVOt77kHsrl+UEivuF6W77qpxpUrwqAhVwlBKZMYMkQceEJkyxT5H iaRevURoA16/frDjU2KjmfGpgRWfpgfqsqwcTH6OO07kpJOCHomSCGow+0Up3qz99isVpcSXqjdL yTYqQoscuhE9+KAtQI07p2ZN69Y9/nh7YlLyAxWhqcH+fcghthWlUnGICx861JZfU/ILzv1+UUoF E0TpEUfYhFKtz6xkAxWhRQg9ke+6S2TiRJGffhJp3lykd2+7aIZlfoIIxVr92GNBjyQ/YN9/4QWR hx4KeiT5DVePRo1s2A6tc5X8heRSxCjJpnjEaEaApRtBSva9dqhTMoGK0CLhrbdsIsakSfYxZZP6 9bPCZb31Ah2akgYQoffeaycSSmJwTVJv8ZtvtENNZRk92rbXve++oEeipBM62GHpRpDiJcNzgCBF mGpolpIuVIQWMPQSv+02kUcesY/pdIL7Ufs+Fx4IKazatWoFPZL8oUkTW+2BBA2l4rDfbbedrV2p rTwLkx9+EHnuOStIuW3QoNRKSt7AuusGPUIlX9HcuAJj7lyRE06wooSYTqycb7xh3Wa4HlWAFh6E V4AK0NTgWKCrl1I5ate25xpKtymFCY0dSFAlf4BY0nHjbGb9WWeJbL65XUdCK61IFSUVVIQWAIsX WwsnwpN2euCE5/332zIqSuFC/VYldWio8OKLQY+iMCCZUeNriwPEJ9eZYcNs6b5Fi6yrfto0ke23 L12HC19REqHu+Dzl229FbrzRtmmktBIZjRdcoJbOYoS4XqxQeiSnBq1lKdGE9QaPwdq1tnaoFvVO HWpQbrWVrR2qHZSKl7//tslNxJI+8YStH+uqrSBQFSUStYTmEYjNO+8UadjQ1jmkODnWB8QHB70K 0OJkzpygR5CfbLKJFZxvvmkf0wd9p51sG1olNapXt/GBNLZQihcy6LkOjRkjsmyZyN13W/c9caNU T8BwQn1SRXGoCM0DqMHHCZ7g70GDRAYMsL2DyV6kg5FS3GAV1/IpyYG1k7g2dyH0u+RdTOPGGwcz tnyHuMDJk4MehZJLEAp2yy32eBs+3Lrod95Z5OCDbZlAvBFKcaMiNEf55ReRq66yHSyYQWK1+fhj W8tt8GDNQlXKsu++QY8gP3j+eZu4x+9Fu0mXnMRFkgvkRhvpb1lRCAkiSU4tXUokGFAOO8w2RWHS fO65IjNn2nCYo4+2k5c1a4IepRIEKkJzDMoqEdhNpjNJRVhtcLeTeUhJGUWJhgqn5KDoNq7B5cvt fUrNUOPSueCx0KhVuWI4oUEZH0WJBcdX5862NfSqVbbtK9e5evVs/CjF8oktVYoDFaE5wF9/iVx7 bWlZJdwVK1faJAkOUEVJhHarSQ4snVhDEaKffy7Svr2tE0rHH2jXLtDh5T1YtZ5+OuhRKPkCHj08 E0xciCFlYkjzg7p1bTMVQmU04bKw0ez4AKG8RcuW9j611gjm5oBUlFRg8rJ6tU1WU5KDcBcEKHV1 qXP55592IeRFPQ4Vhxi/bbYR+e47kWrVgh6Nkq9ghGFiSKMVzm0I0lNOEdl666BHpqQbtYQGAIHa CAcnQKmzRpyMClAlVUhQAxWgqUHyEZ1fSJyg4w8ClK4/KkArB7Hru+8u8tJLQY9EyWdoCzpwoMi8 edayzvURjwVdmqhH+u+/QY9QSRcqQrMIrnXEJ0HZu+1mrQbYoXfZJeiRKfnKu+8GPYL8BSGKa941 c6DgtlJ5OnQQmT496FEohQLXyltvtfGjVGDAiEP86EUXiXz6adCjUyqLitAM89tvVmQiPgnEPu00 Kzwpr4QbUFEqwwcfBD2C/KZmTStESQakDJpSeVSEKpmAwvd4C+kQR0dAOjftt58ts0YyE3W0lfxD Y0IzBGVKcO+5LD9qohHXoijppH9/u2/pUVw5CGtgoqilz9ID3XFooEHxf0XJFLjlH3vM5lMQh3zO OTZ2lMmlkh+oJTTN4B7gYoa7AAFK3BkCQQWokgm0P3N64KKlAjR9HHSQxoUqmYc62oS5vfWWLfNE oiGJceedZ7PtldxHRWiaILkI8ekSG+hyhPg8/PBgx6UUNosX277nipJLEF+rIlTJJpRdI6N+4UKR DTe0ZeuoR0rtbSV3UXd8JWG21ahR6WNi9Jo3D248SnHBxKdVK5F33gl6JIpSCkkknAdxkbKPKkq2 +eMPax297jqRzTaznQY7ddL9MddQS2gFoWQEO7MToIhP5LwKUCXb7Lhj0CNQlLIQjlSnjibOKcGB NZQwOGr/kkk/YoQ9V06YYBvEKLmBitAUYecldszVZXzzTRWfSrDQYUtRcg1aoM6eHfQolGKHLPpj jrE1R0nipM4oScMjR4r8/HPQo1NUhKYA5SAoE7Fmje01jfhs0yboUSnFjlpClVxE40KVXIO2oFRt mDnTxtM3biwydKhtWKEEg4rQJLjiCut6nzPHzp4Qnx07Bj0qRbFwIlVyl3///Vf+97//BT2MrNO2 rU0K0fqNSq7RrJnIfffZZh9ffWXPoVddpZbRINDEpDhQFBeXErRrZ4taK0ouweQIy3z16kGPJDlG jx7tiZO2shttUEp44YUX5JdffpEWLVrI2LFjZdCgQbJ1SZPoJ554QrbYYgvZZ5995NNPP5Wbb75Z 3Clrr732khNPPDG8nUTrg2Lu3LlyySWXyOzZs+Whhx6SH374Qc4888ygh5UVsNJTx5GLvqLkKiQY X3ONbRFKR0MWOqopmUctoVHgos7F3QlQsuxUgCq5xu+/29t8EaCwxju47rjjjjLPjRo1yvsO1eX9 99+XSZMmyXkU+Svh9ddflw9LiqF+8MEHsmDBAuncubNZWlEWwEei9bnAHnvsIQcccEDQw8gae+5p y9UpSi5Dc4WJE2290c8+swnHw4aVnmOVzKEiNIIePUqLVlP4FqMKcaCKkmssXx70CFKnV69eMm3a NFm7dq15/P333xuRedhhh5nHHTp0kPfee0/eoC9fFLCK8lqWXeiHm8T6v//+21hgsYpeffXV8ivt kQRPx8syffp0Ofvss+Xuu+824veff/4x66ZMmWIstPC7dyW69NJLw+/p37+/nHTSSTJjxgzzHNsb M2aM3HXXXdKvpCvFzz//LEOGDJEBAwbIkiVLwuNbtWqVfP755+Y+70E4n3XWWcb6+3OJL/APb9Y7 fPhwOf74483YBlNbJk+hViMJIYqSDyA+773XtgUlq36HHazbXv3FmUNFaAnMgLB+PvKICJ4ydjqK 3ypKrpKPIrRx48ay7bbbykslGSuPP/64sVquV1Jxv2rVqkaAXXjhhVHf/9tvv8knn3xilp+iZBNE W49gRPxddNFFZn337t3N81he+/TpY6yThAK8++678lpJZWuss7fccou5jxv9Y65IHlhazzjjDCMc TznlFPnuu++MCL3iiitk/vz50rVrV/M6BC9hAT179pTbbrstPD7e7wT2Y489Juecc45069bNiPHL LrvMPD9ixAhjMb788svllVdekaZNm1biFw8WtYQq+Qji8/77bSb9uHF4MLTofaaoGvQAcgHcmbjc gRJM668f7HgUJRmWLg16BBUDa+gj3mwPa+XUqVPD4stx3HHHyY033iiPPvpoufcuXLjQiEk44YQT wqIv1vqjjjrKWDuJw0ToInBr165txCj06NHDiFTo1KmTPPPMM0aQru+dBLDQ/vnnn+Y51gHW0i++ +EK++uorY3VdsWKF1K1bVzbeeGMjNtddd11jTUVkIzL5zJtuuskIymjccMMNnlDb03wecaPw2Wef GWG60047mZCCjTbaqKI/deDsvrvNQua8Wq1a0KNRlNSg+g1lGOnE5J1OzKTKO2SlYcOgR1Y4FLUI feIJkS5d7H0soCUGEkXJC/K1NzLCr3nz5jJs2DAjuPaj9pmPKlWqGPc5lsYjjzyyzDqSjXDnxyJy /ddff21EnLO0ckv8KeISatI0voSOHTuahW2QPLXUU/m433HLX0PWggfisJqnpkiswlrpkqDYJgIU sIzyGveZ7vlouHW8nrABwPI5cOBAueeee4xYbt++fcz35zqEMnlaWhYs0HJ2Sv5CmB4tQMeMsR3q +vcXYV6Zx/PDnKFo3fG43p0A5TqiAlTJNygtko9gOcTaiFsbS2aVKH30EKbNmjUzltLKfhbgeoe3 335bNtlkE6lDO58IdthhBzOW8ePHGystAhgx3KBBA9l0002NNXXWrFly3333GTf6BjGCxXktAnTR okXm8YsvvpjSmBG9ZPljtcUKWz2fMs+ioHGhSiHA4Y6z4qOPRL75xlZ+qOTpSZEiFKHEdbhr3vTp GnCs5C9ffmm7geQjJN1Qfolbxzrel1nH94Wuv/56Ywl0Yi9yfSTR1iMqJ0yYIMccc4wceOCB0rdv X7mXzIOSdZECuIs3M0WotmnTRg4//HATk9mlZLaKuEQc77vvvsbNjzh024jcDuIVCyaZ8MSNunH5 X+sfr//+3nvvbZKWRo4caVz1keEK+QYW0Pnzgx6FoqQHuiWSSY/3FAcJ9XB9uYdKihRVndBddxUp qfai4lPJeyilWbWqSEmydd5B7KRzWTvImve7rzk9+QVe5PpI4q0nUYl4UP+2WSKFK8XlSZCKvO/f Tq1ateS///4Lf1a01/H9eA2udjcu/2dGjtU9ZowkUiFySbA6+uijw4lR+QgFwU8+GWt00CNRlPRC kQ/yDq++WrwJrnXRu+o6SnLkqR0lNX780Vo/EaCjR6sAVQqDlSutEM1XIgUoRArISAtjPAGaaL1f gLptR7Os+sVkpLB02+G9/s+K9jq+X7WSbBz3Wv9nRo7VPXZW2969extLMWWl8hkqZXla2hPqQY9E UdILh+w559jku9Wr7b5eUtlNSZKCt4TedZcNIobvv8elFux4FCVdoM969hSZPDnokSjphjAE6oYS 05pIeOcDJCfROSmPq00pSkJefZXybCLkU44apYlLyVDQltBttrEClB0Bqa0CVCk08tkSqsSmRo0a Ur9+/YIQoNC8eWkolKIUKjRDYz+nJBn7vNYWTUzBilCsRCRuUFKhpCSgohQc9eoFPQJFSQy94xcu DHoUipJ56Dl/zz0i9LqgtBN9N0oqwilRKDgRSgFvF0ZGHcXzzw92PIqSSRo0CHoExQXRSxSoV1JD LaFKsdGhg8gHH1hjGGXKPv006BHlJgUlQh96iLaA9j7ud+1qoBQ6m28e9AhykzfffNN0TqJt5tCh Q02Ho3RA8ftDDjkkLdsqJlSEKsXIZpvZUk7nniuyzz62DahSloIRof360Q7Q3i/sVCtFKUXd8eWh 5zv1QOkJf/7558tWW20l42gArQRGo0a2wLeGRinFCOWbnntO5IILRAYN0koRfgoiO36HHei3LLLv viKvvx70aBQlOxB2wkVdMzBLweK5++67ywcffGDEZyS04STz/Pnnnzddm+gbf+edd3rnjddls802 M+0yt9tuO/M6an/OnTtXli9fLn369DGF56nfyS0djejmROH6U089NYBvmn9QtH7sWOuaVJRihHKR GMs4b9OPvn79oEcUPHlvCeVCjAAl9lMFqFJsqAAtyxtvvGF6vyNA//rrL1PsneV76rOJbd+JoMRK ighdTXE/EVOLs3HjxnLaaaeFX0edTlp5UreTrknfYMrz+OKLL+S5556T/v37m65Or1KXRUkI1lDq hSpKsbLJJiLPPCNy+OHinadE3nkn6BEFT/kKy3mES0C6/XaRkmuHohQF334b9Ahyk//973+yMemp QpLiUrn88stl5cqVptzRYxSqFDJWexgLqGPAgAGyePFi0yP+M2a0JVAontfC/vvvL7NnzzZWULY1 FpOeB92MFi5caJ5X4kO8vu/nVZSiBN1CJ96WLUWOOMLWMu/cOehRBUfeWkIfeMDe3nSTClCl+Cgx 7CkR7LjjjjJ//nzTLnOXXXYxwpMEJR47atasGb6PVbNVq1by4IMPGhe+/3X+LkgI27///rvc83RE oj2nkhgVoYpSCgXt6a505pm2sH2xkpcilJkEXQmOPdZmnSlKsaGW0OgcfPDBsuGGG5qMeCco/42T BfDss88a1/zIkSOlY8eOZdbNmTNH/vzzT7MQM4r7Xqk4KkIVpSy77y4yb57I/ffjkREpxvls3olQ 54KfOFHk0UeDHYuiBIWK0OjQYWj69OnGRU6CEUlKZMgfdthhZj292/396HkeN3v79u3lhhtuMALW sXbtWiNQd911V+ncubO5jXw/feAj+9sr0VERqijlITnpjTdIqhTp3dt2Wyom8io7fs89bSDviBEi Q4YEPRpFCY5bbxU5+2wtRxYPLKE//PCDyXp3QpHTHQvi0f86MuZxuSM8EbJkv5O0dOWVV5rnqlev Hn49llXnkue9kcJUiQ2REFxsa9UKeiSKklvguOnZU+TXX2090Q02CHpE2SFvLKGXX24FaNeuKkAV ZYstRJo0CXoUuQ1Cs06dOmUEIvf9AtS9ziUzRfZqJ+bTYhvOBwAAIABJREFUL0DBHxOqltDUoJye do5RlPJwWqLhDqeiTp1E/vgj6BFlh7ywhGKqpgYo5P5oFSXzTJggcuedNp5IST9r1qwxFlAnTpX0 0L27yDHH2J7aiqKUB4sobnmqx82cWRqCWKjkfIkmRKcKUEUpCy4bX5K3kmY20gKsGQFL6PLlQY9C UXIXLKL33YdXxjZ4mDvX3i9Uct4d7zxnKkAVpRQVoUo+suWWIqtWBT0KRclt0D1r14rUrm1LOBUy OS1C6bABH30U7DgUJddYs0akRo2gR6EoqUEn1a+/DnoUipL7IERJUCIc8cYbgx5N5shZdzxJSMuW iXTpIrLTTkGPRlFyC3oPqwhNH19+aUujUEZIyRx169pYN0VREsM5fsYM2+KTc9PRRwc9ovSTs5ZQ yjEBMwFFUcqCO15FaPpYuVKkW7egR1H4qCVUUVKjQQOaaoj06yeyaFHQo0k/OSlCt97a3hZLiQJF SZXffxeJqBykVII99rCF1H/6KeiRFDb16mlMqKKkyq67igwbZjtFFlpXpZwTob/8Yq0SBx9cPMVa FSVViAnV6kHpg9Kf++wj8uqrQY+ksGHitN56Ij//HPRIFCW/OPVUa6AbOjTokaSXnBOhrpPGSy8F Ow5FyWUQoTpJSy8HHijyyitBj6Lw0bhQRakYd99tW5a//nrQI0kfOSVCXbwDNbIURYnNn39SyzLo URQWKkKzg8aFKkrF2HxzkXvvtW558gIKgZwSoc2a2Vt+YEVRYqOW0PTTurXIkiU2JEjJHCpCFaXi tGtnG/jccEPQI0kPOSNCiQOFKVOCHYei5AOUE9Ls+PRCrCKlUF57LeiRFDYqQhWlcowcKXLbbba0 XL6TMyLUZcRrmRRFSQyW0A03DHoUhYe65DOPxoQqSuVAL511lsjgwUGPpPLkjAiFI48MegSKkh8Q E1qtWtCjKDwiReikSbY0ipI+1BKqKJXn4otFZs8WmTcv6JFUjpwQofyYQGcARVESgwjVOqHpp00b myBJRyrqFJ9+usgVV2jN4nSiIlRRKg/nf8o1jRoV9EgqR9ZFKCf4ESNsTJvj+uvtLSf+N9/M9ogU Jf9QS2j6uO4625Vk7lz7m9KtjRIoL75of2cea+hD+lARqijpoWdPkZkzRb79NuiRVJysi9DRo0Uu vbR8ZlffviLHHCOy996aGKAoyaDZ8emBwulffSVyxBEi77xT6pJ3npmjjgp2fIWGxoQqSnqgYQm6 ibChfCXrIrRPH3uLCP3uO5E33rCPDzjAKvo6dUSaNs32qBQl/1BLaHq45hqR7t2tGD38cJH69a0I feYZu15FaHrhHE8ZrL//DnokipL/0EnprrvKPvfww/nTZz7rIhSxSQISJyHc8kcfbZ8fONDejhkj summ2R6VouQfKkLTw7rrijz4oEjnziI//CByySUi8+eLrFhhXcctWwY9wsJDXfKKUnGIUW/f3k6g KStHuJBLUHr2WeumP+ecYMeYLIEkJl17rffB3iePG2dP+oBVtG1bLVSvKMmicYrpg97xjzxiJ8ic i5yVjsdVqgQ7tkIEa7O65BWlYpA/Q2Y8SZOEDe2zj41pB/QV5Eu1oUBE6K67ivTqVTY5iQvqHXcE MRpFyU8orq6kj/XXF5k2TeSww0qfU1d8Zqhdu9QAoShKatC+E08ynHKK1VTk0hDeyO0mm4gMGBDs GJMlsBJNV19tT/oOSg00ahTUaBQl//AfP0p6IMThySdtchIiH++Mkn5IqqPygKIoFeOCC0QOPdRm xj/6qK3o4aygFLLPl456VUIeQX34ueeK3HKLSJMmNogWl5gSHO+/bwOc//nHPibW5OST7X3iTe6+ u/S1ui7YdXfeKdK/f26MpRDX4e467zwbLxr0WApxHeWvuICOHx/8WHRdcuuIk16wwE7QGjcWJQeg 3TlWULwKNWuK/PqrrR/6xRc2ATAfCFT2XXaZyL33ijzwgArQIGAH5oRDXAnwH1CZYJ0S+/h225W+ FvO+P0FD1wW3rkULe5/XBD2WQl2HFYEmGi+9FPxYCnHdhx/a8jK5MBZdl9w6QuawtnHd5hx0zz02 wUwJDmKrMUh07WoFKPTrlz8CFKrMmzcv1Lp1a09F1wxkAGvXNpV11w2mlsCv3r82c+ZMaVuEPreJ E+1FdvhwW+JByR9wY3JBCM6HoSiV44wzRJo1s7dKfvHff/b6wSSiW7egR6MAcaH8J3huli0T2Wab oEeUPFVnzZolgwYN8mY3lwU9lqxz6623ygsvvFB0IvTdd0WuvNJ2p9p++6BHo6TKv/8GPQJFqRwa E5q/4CmjuYySOxDW+OqrIvvtl18CFIwTfF1PPm/s940UCVWLNAbg44+tAFVXSn6iRb6VfEdFaOFA 7PSWW4oMGRL0SILnyiuvlHfffVeqBFDXbautmsiKFb9Lp05fZf2zYYcddpAbbrgh5e9enCqsyDn2 WM2szmfUEqrkO4STUHBbyX+GDbMxosSS0nGsmLnqqqtk2rRpss46gRUeCozOnTvL1VdfLdXJjEoB FaFFxMKFNp6HbDolf+E/VJR8BksoXfOU/IckPpJb6XZY7CK0WrVqcsQRR3j79wZBDyXrbFjB7inF J9eLGDLo/A0ClPzEldBSlHxF3fGFxfHHi+y2m7ZiVVJHRWiR8MknVoDuuWfQI1FSBff7CSeI3Hhj 6WMH8aG9e4uMHRvM2BQlGX76SeT8822dSfCLUM5NbdqIPP54cONTKgdZ2ddfr3kGSuqoCC0SONFT EkXJP37+2fY1HzRI5IMPyrrjEab33y8ydWpw41OURJC5e9NNtqSPKzHmYkIHDrRF0qnaoShKcVEp Ebpy5Ur5N4AsiaA+N5/57jvtNZ6vbLaZyJlnUlPXdhnjFnB9kRQAruGAouQiRxxhJ8GffWb3WWcJ pc/19Om228s55wQ9SqUyLF4s8sILQY9CyTeSEqFDhgzxLoJnmnqiL9JvrYTu3bvLB5hmKshDDz0k t912W8rvq+znFiN9+ohMmxb0KJSKQl3XzTcXefllkWeesc/RbID2kl26aI9zJbdhAkxnF6q34Lal 3zUi9KKL7Hqs/OzfSv6CNXvy5KBHoeQbSYnQ672zxrHHHitt2rSRfv36yaOPPpqWD99jjz3kgAMO SMu2FKWQqV271Op5ww22YDTtbrEojR4d7NgUJRn23lvktNNsYt24cba/NfWK69YVueCCoEenKMHx yy+/yA80gC+Sz/WTtDv+wAMPlK5du8oZZ5whM2bMCD//+eefh62kP//8s6xZs0YuvPBCCZX0FPz+ ++9N7SgYO3as9OzZU8ZQy8Fj1apV5v1AC00KvZ5wwgnGQgq43bHC8p6RI0d6Jy9NC64otPIiLkvJ X+gJvPvuHDc2LpRDDAtSw4ZBj0xRkuPaa23yyqJF1n0LhJJstFGw41KUbPDZZ5/JWWedZRZ00RfM xISWmxNNoffKcJo3w1u0KLUW6On43MqSckzowoULTWV8B18A9zhik9afG3lnE1qBvvHGG2b91KlT ZfXq1Ua40qcdAbrzzjubdQsWLAi/7vjjjzevO//8883zQOcBxO+wYcPktddek/vuu6/SX7hYmTtX ZPz4oEehVAasnzffXPq4QQORwYODG4+ipEqtWqX7MBMpLiWnnhrsmJT00LSpSPv2QY8it3n//fdN KCGeZWqJtm7dOm2WyG7dusnWW2+dlm1lk6RF6NFHHy0tWrSQH3/8Uc4lO6KEW265xQjFU70zCQIV TjrpJHnwwQfN/YcffthYN+vUqSMfffSRN/tdbIq5+vnzzz/l2WefNW7/Vq1ayXXXXRf+zL322ssI 3EaNGplZhKIUM/vvb+u9AhPYFJtTKErgkCHvipqPGEH75GDHo6QHyv/16BH0KHIftNDBBx8sF110 kWy77bbyJjEpQrm9v4326du3r7z99tvmuZtuukmWLFkSfi/rv/76a3n99delT58+xgv9v//9z6xD 4P5GkoDHhx9+KKeffrr079/f6DISuceNGye9evWSs88+O+yBzgWSPvxJINpqq63KdQKg7zzQKeDv kqbWWDV32203ufjii41LfZ999jHPT5o0yfyImKGxijoQtptttlm5/vX333+/3H777dKhQwf58ssv pUmTJhX7looSAHPmzDEhKo5mzZrJNttsY+4zGfOfCFJZ16HDalm+vJnUrPmOrFiRnm361xH7zfGo KN99953MI+OkhBo1aoTj+Am9euWVVyq07rjj3vIuprt6k6h58uqr6dmmW1fVU7Xt2rVL34+gKBmA eEx0TePGjeXTTz81rvGbb77ZnJc7deokX331ldmv77jjDiNGv/32W2P0O++884wFlSTxn376yRjx 4LHHHpODDjpI1ltvPWnfvr0MHz7ciFy8yLVJKvBAe+GVxnX/3HPPBfn1wyQtQvlhnOBMxBZbbOHN ivY01lGUN6DQ9913X3nqqadk11139S6iy8Ov33LLLWXt2rXGUoqrHuXORRLRSpwoJ5TBgweH40yV 1GnUSLwdNOhRFA+Iz/3226+M1Z94aif8uGA+/fTTFV5Xp05VufXWf9O6TWBCePLJJ5sZtKJgbXn+ +eflsMMOM48bNGgQFn5cAG+99dbwa1NZN3XqKKlbdz3vNf+kbZtuHV61Bx54wBhDlOxBjK+nm0T1 f3wIQUQk4tkllwbjGvssFtATTzzRvIbn8QDzGG/w6NGjTUI4oY/rr7++MfpxXJ5yyikmBNLPCy+8 YLzWHLuA1RUGDBhgDA4cN7nkVU5KhMYSn+uss45ZIu8DF7JjjjkmXILpkUcekbvuuktq1qxpXOuo /ypVqpiF93ECQWwS01C9enUTV4qAxaTMn8QYmjdvHvWzlMS0aRPyduYq5j7mfeJu+T132WUX8x8o 6YUkOqyJz7h6ShGwX7Pk+jqluMG6gmUx2jWgfv36MffvINdhYdLrQ/bBYE4JOSdC+R9wFWONq6ox F2EQiPfcc483CatbZj/1/0Z4nLmGYCjgGo3Vk9DGG2+80VyvcccjTDHWcYy2bNky/F4XvuiHBCis q4hftv2fv+NJwCS1Z/zxxx9RT0Ive3uce54flscO4h5w67kkJlQ+We78sLWITvcgNsFZN4877jiT fc+Ou8kmm4TfgxDlj0L5ux/O/7lKfDDVMxFYtmxZeEJAIPSTTz4pV1xxhfk/pk2bJruTdq2kDSZS F2jdGSXP4bybb0SGdSnZ4csvP5AlS/CoXGoe49EkP4SEY6xxiCZ/UnOxwrWhXr16Sb+e35B8GcQl OTNcszfddFOTFM6+zuTLL0KbNm0q9957r4kDRdjyPyBaKYlJlSESvnMpyTup6WIswRf5vP8xVk9n Wnbw4zsBCs4K6n+/E6CODTfcMHwidK9VAZociE3icTH/D6Q3Xgk8RxmsTz75RB5//HETt6ukF/b1 Sy65JOhhpAyxScQSKYqiJAOGJJJsbrmlvZx1Vmm9OK4znE+4vpC5jeGj2InlxXVe4WiPyaQnUcmF lyDqCXfEsulc9P5tt23b1lhIEaOEPrpwmtmzZxtLKOIVXRXtc4OgyogRI0JYH6+lgFsawQyPOzKX 3SLMLgi857YQwdLMzPOcJPrhYWUmCJr4klz+z5TMwoyZ+GtuFYWwqqOOOsp4qfIFSvptvvnmJgZO yTxY2iitiEGDZLFiBlc38cmRCdx+yH+JNKQh5FnctddZMd3rt9tuO3NOJrTBPYeRCY+zE5GR2+V6 zvac4OQaT24O1lP32sjPrQwYXtBT1VMs2ZIxtcFJQMVMsBDOkIwABf4rTPSEUGhTgMpDCAsZjYqS zxDOk2/nA0QAzU+U7IBHjUzrZAUoXtLevXsXbaJxNE9upFfYHx9K0hIxnk6Aum2gsfxWzMjtYlBy AhTYvgtVca+N/NwgUJVYwBCnmwrUHGvYsKGxJCiVg1mo/o6KohQ6JCCnEiJH628sZq4euBIfyi5N nz496GFkjLwUocygVqxYEfQwchKsFtRV/eabbyr0fk4MFLX966+/0jwyJR+gEsXll18e9DAURclx SHRd7HqvpgDWN95L90Q8Rkp8sDAXcphDUiKU/u3+0hgUWPXXa8s2dAw45JBDAvv8XIYaYRzY1F6t CFhCKZpL8LOSOajNeffdd5d5juxHjjWg3W1kgtDQoUPDhYlpBIHlmoV4rHiQNJDsTJoav4e7djZK 0UACCbUJKYjth1rNLvEhGmQ806nFD+cgEiaoRUh3PYpuO5544olwq2bK9rl9mMV13Iscl7/XNo8T QdUPinErmYNqK3h6XAvuVOE6w7muIiJWKSySEqHEtmFCdz1OV61aZU4m8aAMgCuSqmSPt956K1zA uaJgSfXHnyipk6hEE2Lv0ksvNQHiDkplMcECar+RMDdlypTweiYHv//+u7k/atQo09a2c+fO0rFj x5if895775mLPQJWUWKB54N9CnHgF6NUJonnanUdXfywHfZ/xCkilg4vDpIraCkIJLJQLYV9mCVa 2Rr6bC9atMgIYRJdaQDBZ8aDuLdCthzlAq+++qqp612ZzGomz5QcUoKF6gUkQgVFUnVCaQPVo0cP M/PBhB4JgpTyCxRfvfDCC81JC7cuAcskxtCdhRMVpQHYFhdW2kixE3NRxeLDQtvPsWPHmkLqZHWz k1LcnrqgZHVRaoA4R38XmqVLl8qdd95Z4fiSuXPnylVXXWXuc0BRVscFBbNN9+ewDouSW4cFwL8O weHW0XrLv47fwK3DDYHL3B28zPDdSX78+PFl1jGbd+uwmvnXMSlw6zjR83mso0UXGa2VBUscVRMc XAT4L3Rd8uuoThALLAhcdNm3Dz300PD7OX4c7DdYC7hA0yUjEspuJIrFovUtXcc4dl1HsnhgaeK1 /onMkUceaVr2AgkI/pJeuq5w1gHnY0Qk7QHZ/xCkTJhiQR1nOuHhGWNfxJqPyGTfpDseE1p6Y2P9 dO2b/dAww3VjigUZwOyPLJyTGX+8Fs6cC0mydOdKjh1X3oZ1kydPLrOOGtVuHccgsJ5rFeVx3DpE s/99bvLHRJLv6l/H7+rWkVjitsl1wHkaWEf7avc+Ps95+FhHSR3/OndMss5Vr3DbdL8tGdBc09z7 WEeLR7eOUj/+da6+JOu47rpt8l9S3setw0rtfx+W0HTU/KRdMNd2l6REAo7LrubaToKZW8d/7roD oS0417p1nNfcxIPzHB3r3DrKFblzNN+DxDW3jvbiLlmHSRMd49w6fhtXUhKjGkY4t44SSa4VJr8p +z3rEhWBxzuAxuG4ojEPoJ84xqIdH9mAaxhj8tcazSZJiVB+WKw2VO5HNPnBeoPARFzhgqGUBwcW AvPjjz82BzjdLPgTEUgcZJzk+NF5DQcarwOKsnJhRuxx8eTH4QBm57jmmmvCHQIcuIyplVXZLGS3 40Rm62EdcLGRkevYyWOtIx4z1jo6FyC23fN8thMS/A6x1vEbxFrHQcDnsY6i89Rkqwz8T3R08Fsm OOG6Axnrmj/mVNclXhcNLuBYKRGhnMRwTTlBCrRr4zmOrWgud07E7AOctN0JzQ8XKy6cWEE5odPK MFENUMaPlcOfncnJ0QkVrFJ+F5quK5x1fjjXOKslE95YJZrofIfX5KWXXjJikv2NSROiCdiP2OeY XDk3vB8s/+zHCJztt98+6qQKzxtGjvnz5xvRmkj8MIkiJGDHHXc0jymX40QoxwTd+Nx5lHVOhDox Cawns9iJUNZhPPG/z4lQjAOIXv86J0JZ5yzFrOdYdSKUdVy73Pv4PCdCWUdhcf86J0L5bwg58G/T 9QHnOsD1073PdR9065hY+Ne5BjOs439220TsOaHLuhNOOKHM+zD8pFqKJxLCNDCYUM7RCVyKrDsx xH6DbnDruCa5pioYtDjfunVk3LtEXIxACGq3jt8fsQm33367uZa6dZxbndjG0IPl3a1jIua6NKJH OBbcOjoXORHKOdyJ9EQZ/847wEQKvQT8zuzT8UQoxgi6JxVkZzvqhA4ePDgUD2+HNLfeHx/yhGHI m9mGvIulea5Lly4h74AIPf300yFPkIa8k0/I22lD8+bNC+2///7hbXjCMzRw4MCQd7ENeTPnkPeD hv7444/QGWecEfIO4JA3Aw95M5mQd4CZ13Pr7eQhbyYU8g7U0JlnnhnelvcHhrwTZ8jbIUPPP/98 3LHHw5tVh7wZSYXfX6jo71J52J+9E1Xc13z55ZehOnXqmH3dO4mGzj777PA678IYmj59esi7QIe2 3HLLkDcLDzVo0MAcP7DOOuuEvItgyJuEhc477zzznDejDnlC0xyHwLHhCQJz3xMfIU8shLzJS9wx eRMQc3wqxYV3geXqaRZv8hPyRI15vkePHqHJkyfHfS/7ed++fc39du3ahbxJjLnvXWRDvXv3Nvtc 69atQ1OnTjXXAE8UmPXs461atTL7MIsnNkOe6DD7MIsnTsw2dtlll9CVV14Z8i7AoUaNGoU8gRF3 PHwG1yolt/FEfahbt25BDyOtVKtWzeiaWLA/9+rVK+RNuEJz5swxz/mPCd6LJjvxxBND48ePN8fO 7NmzQ54wD3mTEHNd9oRs+PVw8cUXm2MF2Ca/Kyxbtix0wQUXhE466aTQgw8+GH796NGjQ97kw4zD E88hT/yGvAleyJv4hC6//PLQxx9/XKHv7k2UQmvWrEn5fSllx+NexOr2zjvvhBU/pm2sgiQrYZlk puFmwX6YNc6YMcPMmKnoz6yOWRgzVoohsx1mX+693DLTcokYkRYlZptYf4iBVErBqkCf2crAbxqU ab5QSKZEE/svXS2wVOAGxDIaCTNkLDVYRSLjr/BC4MphVg5YFLBkUv4EmL1zvOI6wsvAMYqVW1Gi geUdixru3FSS0wjVwlLJ+R93I3Gbfthv8WIR6hQZe0YnN/ZhFiyyuF/Zh1lcfVI8cFjzqdqB5QzX oRIsXPPx6lUGrIKc/4oNv3cgEvZv1pN46rzMhBrstNNOJn4W7zD1QTke8IYSzkCIikscJ6SP59FT lHbCuopuwzqMJRnIN8CySkdLZ81Fz5166qnm+IsX6pIJknLHOwhSdz8ePwpgyiZg3JmJcVtw0uG1 xFc4uJjyPD8SMWq4SjgxNWjQwPRBdaKWHRPTORdPWngSDxQNCqxiuucCy8kzUVxRscBvScF5SljF cwXHgjgbJgoTJkzIwOiUSBCeXFw5mbAfRwPXG26hROVMnDsNEMGEsuCGd/uBN8M2wjRe4pqWaCpO uNAhPisCuQC4Qon9x20fLVkFYUooFZMt50qOhov9jMXy5cvN9UYJFv5jrtNcZ1xMZaow8cCNXoxg WMB44FzyQJgF4SBcE9jP+X0x0qGtCGnEHU//d6CDEhqJ8AESVBGhiEqu3UzSmBRyzLlwEkQnIRSn nHKKeUyHzAMPPDD82RyTGEWcQSObJGUJ9ceHMevlQuUq8fPlSBhCdSME+/TpY55HZDLrxaLmTm5d unQxPxwiiZk2ZWp4DtipET4UvuXHodsPItOti9ZXFZH68MMPm9c660+xg2gnGYCYmYqAeCEOh99W yTwuhjqyDI6/xzAXXWKyCeB3noJECUmUZOI4IrDebYuTG7Nrf0Z+JFqiqTghHo4JPYt/ApqoRJOD mEtEhYu9hMg+2VR7YB92LQ1j9dH2g3eMCzEWM64pGDZcGbNYxCvRhJfHfc94iYNKfBCeiB/iJisC XpnLLrtM9t577zSPLD+I5h1wSa3EyvL7IDRjdTx0nmW8aJQuQ2Oxb3Oc8N9wnPgna1wH/EaMSAMV RkKu+4HUB08mJtT7kco9Fxlb5v2AJn4z8jXEsvkh7iDafT/eD1huO96FM+XtJKJQYx+JD6lbt244 5iRZIv8/JTYjR44MNWzY0CzE7kRCbMzw4cOT2paLg/aT6JiLtj7ytZHHDKR6rHgCOfw9+/fvn9J7 lfxh6NCh4ZhQ4i8rQjL7cSr7cCbwBG34ex522GFZ//xCghjezTff3MQTpkKiuPR8JpmYUOKkHeTU 1KtXLxzjufXWW4cWL14cXu+OkUsvvTTkic3w88RytmjRIrTrrruGt9O2bdvQuHHjzGNiPps3b27y c+Ciiy4KeYLX3HcxoA73+NRTTw15k7cKf/eMxoTG6nXqB2UdWZvNWSv9+K2q/vt+XJyCfzuRM+Zk tlOsYA3FnE+2aSjJ/rxYpcnwVJKDch38xizE30RCPDOz3GSIFkOd6JhLZAmN1RM40bFCiSZ/Bj0l e9z31AYGSjyS2Y9T2YeV3IYYXioWUNkjUWkiB+dK3MLFSiLvAOEJeIeprkDpswcffNA8j3eKSgrk 0wAWT7yWLncD7yeueFexAS8Y3mnCuIgLJXufqgnRxuAeY9XGmuoPEcgGqt4KFHZqYrXg6aefNjtx tFqTJH4RQ0LZLEIllOKGhBBiSXH/KwqJEiSOxirRlIuQEEjyxoABA4IeSsFDTgcLLFmyJJwrEg3y PQi3CyLuMFdAYPqbi1DiDGHuJmft27c3vyPhhbjQ3cRu//33N2Wd/GKfeqhOTBLrSUyofyKIqCTO lAQ/V18VSAj3TwD9j0k6zzZ52TteSQ2sW1tuWV969jzFCE6sW25nJqGA2Q8ClMBnJT0Qf1PZ+rWK EjRMUl2Wer7ARZ0sXyW7EBNMciXihxhFf3wh9cVpMkNCTLwOb8VAPO+AA29mpGcBw5K/Nivb8b83 micCw5NfgEb7fP/jZOK0042K0CKAbk5XXvmOLFu2lzlRUPKEoHKgWQDmdxWg6SWZEk2KoiiFwuDB g01SGM0QSFBDbLrsd6zphPQkk+SmFBfqji8S2rTZ1ptF9fdmTuuYemDOuhGtW4qSHogbxRrqemdj gfa7opih5to6LdGk+PHX9aQ2JO05HcT7U5Yp19aRZexCkZTsQSOuqlWPlDvuONJ0WCKekbI/XG/8 rbYVxU9SIpTCqQTPAubdTMZ0UEaGwtxqmUsvlKDTyjdhAAAgAElEQVRkeftt276MVniUcdBSTJkB dwrlkIjb8YP4c0S6PXJhnZZoUvzQfMSVevnkE5Y6nrBYYR67ZAogvMe/PwW5jutHZHKrknnmzSO+ UKRdOwmXTIxV51tRHEmJUJJWqElF7EC6M9Eje6ISXEtQuZIZcMW7Hrf01z344IODHlLBQg9tFgfx PLF6t+fSOkVxICIosEE5yMGDa8lVVw2QQYPKv47JbKz9KdvraNihBIuKUCVZklaUFKL3B7BSJJXM K0oJAC48BCWFsIkDobAwsYaUB8AcDwSLU6CV1m7EiOAKphAxpZ2cxQiB5Nq+EUOCy4USEO286ZVr azhmzBhTqoBSDxTNp+gtmWTffvutaWFFhi+f6e8IUOwsW0YRXDqSbGQK09Lek24LKkIVRfFDhizt AuH772vKffcd5J0v6pvHL7ywQGrWfDPA0VUO4hWVzEMLYVCDUllIEqaFpiudSCIXhrdMwDGMtiI0 IpdJWoQiWhChWFCI86A+GO4QJ0InTpwo5557rhGmCFJKBiAE6VmNIKUVG8HK9erVk/PPP1+mTJki hxxyiCnpsOWWW4bbS/Xs2dMk0iBmqXOF4KRzAPE+ZGqyXXqf0t7qqquuMl2VEKFkfVMHiw4MdML4 +uuvM/OL5Sk0raK9LB3xaLXK/0nJDKVi+OuvUp8t3zKIkyUynEApfOi2xSJCG9gx3kJ3Ffb3KjJr 1gJvOT3Q8Sm5CW3gXVSEWkKjg+EHDyS934F2nOmCjkcDBw40ycaAMS9aWcZcI2kRSuYbVsqGDRsm jAll9oMoBLKwKZTaunVrU38QqyZtpVq1amXWu56oFGYF1w6U2pbxep/SlpLWa/zIrig48Wwvvvii sZoWazuwZKAnLVZqDgilYvhLY9C2lkVR8pWyF8N63kLrTpdMMstbdvcWYkMLx7KFMUVJH97l2CyU ZnLxuipCy4NOwbPsIDfjtttuM8Y0IJcAwxyaKJbXFyZPnizPPfecMdJRExcvMMYlf6vPbt26mWRT jINoMgQwnlAMgTQYokYoBhQELO+l3S26Cnj9G2+8YR5jGMwUSYtQfhi/O56LcKwuCf640WrVqpkv 6XqZIkCTIVHvUzcWts8PDNQgowcyrn6KLA+KFrykGEso4JriP8x2XbBCgG5UxUajRo2CHoKSIThf klX+/vvfypQpl3tCooZ3bl0j++//sCxd2sq7YL3rTVoP8y6gTeWYY6L3Zc8niBvlQqykH8LhHOqO Lw9J3gg+QJBS1xZB6UTo66+/boQiIjSW1/eBBx4wxkDCG5cuXWpCFzG+ffzxx3LccceZ7RAWSQ1w Ct3jkcYLjbf4/vvvN6WyMAriDUU3YdTDGNipUyfTu551M2fOlPHjxxvhmkkqnGXEDOe9994z91HR WDjjgbLHxY57b+eddzbW0WbNmhlLJoIzEqx1EyZMMAKT16D4iS+NByK1f//+xvXPj6witBT0w0EH 2ftOhPJ7sdPyfyip0a9fPzPZ4uAvBjjeXWy3UniwL+PKI4pp1iyR1q2JKdvIOzf3NecNvPO4W6tX 385kzCtKJITbkndQp8534efUEloefwtN4jWdVzgW0by+U6dONWGHhCweVHJhxyCHkc95lRGSwHWe 8BpKBlLQnoQ+qke4ikfEpLq4VDQTWo7/Da2GoSrT5bWSEqHRevyS9UvWPG52FDQmYayjbnFgZXN9 rEkyQq3jBiG2lOQmysGwLXqX89hV7Pf3PsUiSjYkswW3TWe989/v06ePKRLOTIw/SCnFlWiCbbfd 1pjiSRTDJa8iNHU4Jvr27Rv0MBQlrVA2eNWq0sfM4zEYlnRmlJJ8E0UphyvR1KuXFaFMbJL1fBYT JCPhWXZgyIjlVYZoXl+EYrKeKSyt1IF2HZW4RX+RYwN+zzWlzvBct2nTRiZNmmRc/FdffbWximYq vjQpEYqSjhSiuDNQyQgZBA1WTvcafzLDNddcExalmInpQYzbx9WnjOyJyizBbSeZ3qe0neQxEOfI n0NGGH+YEh3+DyzNmP0Roc58ryiK4sDIjzHlttvEO8dzEaQTGNcDYveDHp2Sq7ikJFzx0VpSKmXB yLZ69WoTR0vC65w5c4w7Ph5NmzY1ugf9ROMGFjRPNK+ya9yA653rPi260V/xrNRYSbGoEgqA5xSh TIJ5JkgqGDCaJdSBAI18jV9ZO0uof1uRBdL9PVEjPyvV3qf8oSpAy0OJpldfLX3sXPKaIV98vPWW yLnnWkGhKLG4/nqR004TccasQw8V2WknzsnBjkvJbVx5JnXFlydab3asxbjlEYiHegeZ8ypHvt5/ n0TtRx55xFhVEYu42hGJdDhr2bKlzJ07N6y9WAhtPOaYY4yHGQ8e8aUQ6bl2j922SaDC4tq4ceOM /SZVRowYEcIySXBqsXG9d5Zl1sZtofP44yIPPoi12D4m246MugYNGsiKFSuCHZySVbBuUVmNpmR3 3y3Stm3QI1JyDWJDiQElf0K1hJIMlAH87DP2mctl2LBhpgY15euKCQxquL/9Hbwi8XuN/WB9xOCG NdSJwcjXRj5GfBLf6cQp7+XzMfThXY4UmazzdxPj9Szu/YhYZ0TEMosX2mXjJwJDInrKGRSTRdOi iwQuJL6kxbAl9MsvvzSmfKV48M5DBuYeVArp16/0OUWB0aNFTjpJBaiSPFRFJL/F745XyhPLs0wY ocufccIxntcXCIv0W1Z5n/M0R3qhIbKdrfs8h9+LjZhMVoBWBhWhRcJuu4lQacFVuSKmxO2gmS7B oOQWLmyIcw+uVayhWL2eeirYcSm5wQ8/iNxzj8iFFwY9EiUfUXe8kgpG9mLixSVfbGB6LhaI6xo1 ylq/iC8mlpdYD1qokpykLU6LB2f1pHsWnhMM4XPmiHTqJNKjB6V5qMsb7BiV4Lj5ZpGuXSlgH/RI lHyB+PKbbrKx5totSUmFqlTjp8yS6xWcTYg3cDEOrnxANiGzn971xcKAAWUf45J3IlQpHpwIJdHk lltE6PJG6NaQISIPPyzSsqUtzaMUH5QOJBue5DVFSZbLLxehwE2NGuqOV1KjKkVQ/X2wswktoqjI T2vOhx56KJAxFDOIUOqVaYZ8ceHc8VtsITJwoMiwYSIcfkcfLXLffRRQDnZ8SnCMGydy+OG2uYWi JAM2jAceEHn7bftY3fFKKlS4Y5KSn1CqqWdPEdrLUhICtH1nceEsocSvH3+8SMOG1PYVoWfBFVcE OzYlOKhdfeONIjNnBj0SJV/A6HneebT1pgmKfc7VqixGEfrXX38Zz3K2Pbtcv/HsAglO8cpqZgrq yfsTm5JFRWiRQcvziROxghL/ZzPk6TKFW75JkyYBj07JBiXd2owIpQQvCShXXWVd8UrxQjIS7Tqb NQt6JEq+gM70V2FCgBJmZ9cVnwil6Dy917MNTXroPgk07aHWaLahKH5FuiqpCC1CdtmFeFjcrw2l WrWNvNnbGhMXqiI0t+ndu7fMnj1bxo4da1rdVhQ62mK1KGkxLGeeaScn9H5m31CKD3I0r7uOntRB j0TJZdCX1BkeP15kxozy6//n6+tajDGhW265pVmyzTe++otUvtmJrhJ5gorQIoW2e7NmVZFWrXaV d9+da+JCJ0zoaly1zpI/eXKpi4U4Qdr4OXRd9tctW7bMzLLPOutNueGGzpXepmu9SMIa98mWp3JC Ln73INa1by9Fczx8+63I1ltbS2g2x1K37l/Gddm79zo5+bvoutJ1HTvaJEYSF+mkFQ2XlATUsFSU RFQJBZWVJKWJST169NDEpIA4/fTTTWWEjt4ZZty4J2XVKpG///xPqi9fJDu33FA2aGbbdb33Hh0U St/HiciJGF2XnXUDBw6UMWPGSIsWB8qtt76c1s/7/nuRbt1Exo4VOfbY3Pvu2V7HWRHP0vPP257p uTrOdKwjq5nY4DvvtMI7m2OZOHGcaUG4//6dvNv7cup30XVl19ENqV49W9bNgFk0IvaRXuOdOnUy cYkuRlHJPFhCXY/4jz76KK8soWTGB8aAAQMQwCFPhAY5jKJm3Lhx5j/Ybrvt7BOffx4K7bQT12C7 HHlkKLRmTbCDVAxTpkwx/9VGG20U+vfff9O+/euvD4W6d0/7ZvMSDoOttw56FNlh6tRQaK+9gvns E0880ezTXbt2DWYASsXp3TsU+uabMk9NmDDB/J8NGzaM/96XXw6F+M8PPzwUGj8+g4MsDlavXm1+ dxZPhAY9nJTQdOgix7Xv/OKLL+zMlSrVS5aUvuCZZ0Quuiig0Sl+9tprL3NLItnChQvTvn1iQ3G3 ERta7GD1adw46FFkh+HDRS69NJjPfvPNN83t3vR8VPKHBx+09dwef7zM00kVqn/1VZG2bcmgsa4G 4oGoFacUJSpCi5xmJamw3oREFr32msg775R/EScKJXAaNGggW221lbnvLt7pBDcbRerJlC92ikWE kmSydq1Ihw7Z/2wyej/99FNz302wlDyAoFEXFBrR6zcpEXrNNTYTzg9dM9R9X5SoCC1yatWqJQ0p FOmxgCuvCwDyU4SlNnIVd7HOhAgFtYZaikWEjhghcsklwXy224dJTGpJ8KGS+yAee/UqrfP24oul 9yXJbkm+DPoy23UFjJWiQkWoEnbJv7dokcjFF5ddSWpwUL46pRyZFqFqDbVgoNthh6BHkVlwfHz9 tUj37sF8vtuHd999d9lggw2CGYSSGpwY/Oeev/6y5vQSkuqW1K5d+efwyFGeQSk6VIQWOs89V971 EYHrnGR6yA8dKnLXXbaxOB2UcLcE4atTouJE6Mcffyw///xzRj5DraHFYQklFnTIEHuYB8HcuXPN rbri8wiuFVwjODjIxqZ0hC8uNCl3/JVX2hIcDjxxU6aIVKmSoUEruYyK0EKH+JvBg+O+xFlCSXYh NlT69aOIqEjz5vhVsjFKJUlatWplWrLxP7311lsZ+QysoeSica0oRjgEli4tbBH67rsc7zRACObz aTP4dkmzcRWheQSJq5wY6MhD6701a0Suvz68Oil3PCcYkpKOO07ksstEPvnE9gxWihIVoYUMMTZY G0aPFnniiZgvcyL0t99+8y6+S0tX7LmnSMmFQskNNtpoI29u0Nzcz5RLHs44Q+Tll4vTGrpypW1p Gq6HWIBgBWWikeUW12EWezvWL7/8Yu5rZnwe4tqrEa7lc6Mn5Y53UGe0RQuRCvQbVwoHFaGFzMyZ 1qwDffpY804UGjVq5F1w7RXXuOQdKkKzRwo9IyocFzpvnu3Zib/9jTfivrSYraGFHg9KBTbiQU89 NUMf8OOPtvtBHJwrnhaHQfS5VioBkwdCgRo0KPP0v//+G55YJCVCsZpq0mvRoyK0kCEe1MFJg5Y4 f/5Z7mXrrLNO2LqmIjQAOBmnUCfPidB5iMpkIXlgv/1Ebr1VZNw426Nz2rS4b8Ea+sorIuSrFROF Hg967bUi550XvRBGpXHtlxKUdXMTKHXF5yEffSRCR56IGM6U+8arCFVERWhh478QEECO64SSGlFw Lnl6yIdBmC5bZuN+lMxB/O2NN9oC0EngLtz+OosJIeUd95eD4pAJhG+xZsoXsgj9/HOR6dNFzjor Qx9A0goTnoj6kZFokfo8BhEaJYbTL0LVEqoki4rQQuXbb63l8957rQXsoYesVfOoo6K+vEyGvINY HZ6fPz8bIy5eSEWHvn1FFixI+PImTZpI7dq1zf2kXfJffFH+uRUrEoYBFKM1tJBFKDkk1BnfeOMM bPzJJ0WGDbP344R7UNWB/tYQ0xLKZPjII21cRM+eIqtWpXu0SkWJIUJdUhKetU0Iqk4EIRubbZbu 0Sl5horQQmWLLax1jfRXXK9+cRkFZwldvny5SVAKoy75zOOs09Tc69IlYTxdlSpVpE2bNuZ+0iJ0 n33KP4cVKkFZlGKMDS3UmNDVq0Uefti64tMOGc7+VPsvv7Q/ZBSo6kB1B6o87LHHHuVfwOTokEOs RZUZAYPmHPb77xkYuJIyCUTopptuaoRoXEiarVFDk5IUFaFFwe67i7z3XtyX+Nt3fvjhh6UrVIRm Fi64/oQx/KVYfnCXx8FZkJIu0zR+PBlopY/JaJ00Kam3nn66TWTJQLv6nKRQLaEUyUAnpr3qGtb0 Cy+0CSt+CDOJgps4MfGl2kM58N788EPZ5zhGZsxIx2iVypLAHa/xoEoqqAgtBpIQobhPttlmG3O/ TFwolopo/eSV9OCsoFgFata03akI2Pv777hvcyKU8Ik//vgj8eeQgYyKxLrEfcIzklRaxRQbitcX VzV/RyFBwvrdd1utmHawphMDigUfizstmI4+OmYMR8Ii9bEsnmoJDR68NVi5o5w7kipUX/piFaGK QUVoMbD99tb9EWldiCBqXChZkMSXJnivUkEQ+R9/bPsp09mK4s0dOyZMXW7durW5/eeff2R+sjG7 tEZs3952w/Jbu5OgWKyhhWoFvflm26Qmo50RN93UWvZJv0eUUokhAjwtrqpDTBHauXP5MBEmaNHa PSrZhXMVHpUobnQVoUpFUBFaDHBCpyhwAmuoiwstI0J5b6tWag3NFFQgaNLECkSKPyf4jxzEXZGg BCnXC61AiEWxxIYWYjwoId633SZy8cUZ/iCCTrFWMumNwSeffOLNZ+2ENqYIJd75gQdKuwVQjxJR u9VW6R6xkioxXPGg7nilIuS2CMVa88wz1hKnVI4kXPJOhH4YaSXTuNDsgHUzhdqfKceFOioYYlGI 1lCSuM891wo1KERL6B13iLRtm4Xvxb5bYqGPhZsw+SdRUenVywawkm1PZYeDDkrjQJUKE0eEqiVU qQi5KUJJyiA5A1FESSFiFe+7L+hR5TdJiFDnjqfrBVnyYVSEZocKitCULaEcV6itFGPsiBDAmlZI 1tAJE0RuuUXk9tvtY78IZf6LQCUMLl9h7CQkXXJJFj6McwTnijgkjAf1w5/BhClBBQcli6RLhGp5 JqWE3BSh99xjy3I4OJP272/dPUrFSEKENvauvhvgFhbtnBQIKYpQV6bpq6++klWp1FGkYThhAO++ m+oIjTX09ddTDinNWYiTBIoHkOTtRCjlWrt2tQKV6kP5CqdSDt+ShmiZJQkRmlKnJOIPd9wxHSNT 0kUSIlTd8Uoq5KYIffXV8s8hRFUIVRziDSn/E8f6Rd0+V6qpTIY82dT8/l9/ndkxFjskgX3zjU1l TgLCJzYsSWByFqakqeDEgo8jNrRQMuWPOEJk221tBaCZM21MKAaaY44RoehAv35ZEnAZgDy3666z BReyQgJ3/Jo1a2RhSSxHwk5J/PjEGJZU7FByAFqyMiPjPBWF70vqG6s7XkmF3BShBKJHI6OpnQUO 2YwI0QRF66NmyEOKVjqlAlDgGfdjkr9zVe8/dcW+sxUXCoVkDeUnx8kCN91kRTYdhYhGYZePkuCd N0yebBOZSwzmmQUVT12rLbeM+RKy4teuXWsKmbdOEDtqZgMMPlHRcyV70MKZ5LAolTt+/fVX+fPP P819FaFKKuTmEU6dxMiTGaYJXMpKxUkhOamcCFWXfHbIVlxoJf7PQosNpVsqEQovvGA1z/PP24Zj jz0mUq1a0KOrGIQWUCkpK7GgkIIrfuedd5aNE/UNxeIWL3FJyT5JuOJB3fFKKuSmCK1Xz/YrP/NM exWgvog/RlSpGCmI0KVLl8rvfte9itDsUMG4UGqF/ov/NVm4mBBH+vPPqY7QgLVwzpzCsIYy36Vb KvmQFOLAafDII/nteJk2TaRWLVsSNiukIEITuuJB40Fzj8WLrTctCq48E2hikpIKuSlCoX59kWuu sWaXM86wpgqlciQhQpuXBMD9999/ZUs1qTs+O1TQEsqEoZz1Oh6Y/Kj/qtZQAyEGDuIo870i0IgR WbSCQgrlmZJKSlJLaO6RhCW0WrVq0Vux+sFMT61YFaGK5LIIBYTnP/8EPYrCgXhPWunFsZht5p0Y ti4xAZURNfgn6Vri73OupB9++3XXtbURk6C+N1lz/1c240IBayh1NlPRvrkKopN8ix49RC64IOjR VA7CCej6SufMrIAJmcltSXxyNJYtWybfltR7VhGapyQhQrfgOpEIEi8Jx+A8p1QMrsMvvRT0KNJC bovQ9ddP2ENbSQE6kJAKzMkkDs4lXyZDHtQlnx3yIC4UCi1TftQoWzc03xk+3FpBs1ZeEzctE6E4 cZ5u3yQWdOcYQqYMKkJzjyTc8ZqUlCWuuMJ6iguA3BehaglNLykUrdfkpICoYFxotkUo4MYmNrQQ rKH0xUjkScx1+C9WrhTp3j2LH5qEK96VEGNfXSdRxjsiBZdtMgkuSnZgp+LgqF076mrtlpRFONlS +uLll3F/BT2aSpPbIhQw2aeScKHER9t35j4VtIR++umn8mOSNUYNlMD59ddKtcWlt8HgwYUTG5rv YAXl/8iqpzPdRerVCpp7xHHFg3ZLyiL+YG9KYOQ5uS9C1SWfXlIQoT/99JOsWLGidAUxX3TZoWix kjm4oFMdIsnfuVWrVqZmaCgUSj0uNA0TC2JDMXRFRm8o2YXDGiPJSSdl+YMTiFDqR7rQnqQy41WE 5h4JRKhzx2t5pgyDq2PGjNLHTz1lwyTymNwXoZqclF5atrRXK9xdMWjiXQCqlRRILBMXSs0Xymcl iClVKgm/MzF2JJElAV2TXAhFEC55rKFkyhdKbGi+Qkb8hRfaeXvWoJMaF8E4NZzfeecd7xRuz+Ft kqmcryI090inJVRFaMUZMqTsY67jI0cGM5Y0oSK02Nh0UytyaAkTA6xqTZs2Nfc1LjQg8iguFNQa GixLlthux677U9ZgQkuySpyq/m6fZHK7KeefRGiN0NwjTlISqAjNElhB+f044SJI6Sx22WVBj6pS 5L4IVXd8+tHOSblPBeNCaY0YimPlLkea/k+NDQ0WjCHnnGMLYGSVdMeDQp5bQkePHl2ussgLL7wg jz76qHz22Wdy7rnnyldffRVe98QTT8gb1DoTG9d91llnyZlnnmmW+++/P6tjj4m643MDyiQST4su ostG48Z5faxA7otQtYSmnxREqJZpCogKilASkz7hIp4sNIUgi8Uf+1tBBgxAcKg1NNvw1xEaRrfj rJNCkfqk4kGZQHlCTXbYIR2jC4Q1a9bIHXfcUea5UaNGeROE6uZ8OmnSJDnvvPPC615//fVwEiiT /gULFkjnzp3NQrx34JDs+McfNhQrCjQ2IX8A1BKaJTDMZTXuJnOoCC1GUijTxMydxIIwxJQuXKjW 6UzTooV1S3LyT4IdvIu2c3UG5ZJXa2gw0OEJ7xxRNlkngSX0yy+/lJWU95EkLaEoagRK1k266aNX r14ybdo0WUsRfyEZ/HsjMg877DDzuEOHDt7p972w9TMSCr7zWpZd4rjAs0YCVzzf77+SJErNjs8S KkKziLrj008K7Ts5kS5EdDqoUI6VohAKQ+Yy7PfNmtlqBEkSdFwoYA0lQX/BgrRsLm1MnTpVhlO/ yMfQoUPDE6yLL7447AI9//zzjTULt+h/ERUKRowYIZ9//nm2hp2Qb76xJQO9IWefX34Rwa0cR6C4 fZFWju6cEpcCiAdt3LixbLvttvJSSUebxx9/3Fg11ytpPU3MPfvihWSRReG3334z3gwWZ2EMlCST kkDd8VlCRWgWUUto+tlmG5vVyhUsBpxMttpqK3NfXfIBkSedk/zkaqb8Y489Jtdff71MmTIl/NzE iRPl999/N/dxlx599NFGLHTs2NGIpldeecUsDi62N910k9SL4ZYMgjFjRE48MaC67rR8ZUIbpyip 2xf39PaxdZMpXprn8aAOrKGPPPKIuc8EiMd+jjvuODPBJ040Eib9F110kVlmzZqVlfHGJcl4UEgq 8UxFaOVREZpFVIRmBu2clPtUUIRyEXPiKimo/0pd0lQSmuKQq9bQc845R4YMGeKdv6N7VpwL9OCD DzaP/UICSCDp1KmTd+7PjZM/oXp33WVbpwZCCp2SiiUpydGjRw958sknZfXq1Sakab/99iuzvkqV KiaB6ZJLLpF/I5qx8Fuxr7F07do1m8OOTpKW0Nq1a4etvTHBs4B1NxmxqsQGTZTot84Tcl+Eqjs+ M2iGfO6Toght7b2eixsXNWozJg1WiU02seU+0kCuxoZycW/RooXcdtttUdc7F6jLXO7Zs6eJ7XMi IZpFK0jGjhXp0sWWlA2EBPGgiP13S8JJik2E1q1b1+xrZ5xxhhGSHJeRIEybNWtm9qucJkkRmpQr /ocfbOvPRK1blfioJTSLqCU0M1RGhPI8GaypWNuU1CE2jhM8gfxJgCVix5J4uiBd8kC9SvRzrllD R44cKdddd13U9qaDPeWMCxQLFWy33XYm4evFF180yRdLliyRAw88MNtDjsqaNVaEIvYDI4EIJcv7 L8J+JAURWgAxoY7jjz/eWDO5dazjia91fAKMEBFiQDdg5hZlfeBwjl+9WmT77WO+xLnjNSkpi6gI zSIqQjNDCu74H7zZq7+unVStykrrwlUyB9YTLvJ5FhcKuWoNRVQSj4cYjbROYfVENNx4443h55xL nuePPfbYnBEIVAA69NCylYzQb94QTbZ8xiGe3BNP0qhRzJc4V3zDhg1lS2oaJgLB+vXXqP80DTJY +vTpY0S4m8xDly5dZMKECeHHJDGRfHRSSa9V1vtDQAKHLghYpuPs91qoPgBUhGYRFaGZAWvDqlUi v/4a5yU7huPf1CUfEBWMC025hzxxoam48JMgV62hV1xxhSkC/gvZ3Qno3r27PP300/LAAw/kjCse rYax9pJLSh8PHWodFJ6OlmeeycIgUqgPmrQVFO+KJ1jjJTrlG9FiJCMTtCInQ0klcGWLBK54SMkd ryI0PRSQCK0a9AASojGhmYGZLVctFML++0d9CSfQnb0TENnxLEceeWTpSkTos89mabBFDBd6sk+S xF3wV3kTDGo0NmjQILk3IkKxjFPbME0XQY9w7B8AACAASURBVL819Ikn0rLJCuN3c2622WZy6aWX moLhTiTEuvBzYSXWdunSpd5PtEfWxhuPiRNFqGHO4UsVIGqEEs6LlunXDxdvFgaRQqekWEXqEc8s VMky919ZKX/WbSt/zfc991f0+5VdX6OGNfIpCUhChKbkjlcRmh4KKDEp90WoWkIzh3PJxxChgEse ARrVEnr11RkeoCLU/kRZJAmJDpQXos4lIiBpEUo7OEp3LVpk1U2aIFP+2mvtXIf6+0GB5dMvNKkH StKIs0L98ccfMYUoltCUWqFmEOYItOi8/XaR3r35Xvb5pk2tiz4iCbscnErTIeL+vL+5/NWslfx1 WvT1v/zyt3z+OeWHqnmiuLHccEP5bWFfoOU8kxVuq/2+p2xQZVep1t/3XLXo9/235LnEWh/rfh7X ws8uiNDjjov7EnXHB4BaQrOIitDMgQiN0bXDETM5aaedbFwYCR5kViuZgVqtXDmXL7euygQgpLDY Ud8SEdqtW7fkP8uFWKRRhHLBv+Ya6z4Osg12NIHpd4PGc4HmShwo4G7ncCPXxeVWERdKyCXu+Xgi klu+ZjIiLdH6Ot8skmqnHC7VNo++ft68N+WddwZ618mQPP/8XGN5jNxWuYTxUwaK7LNPSpMuJcOk 4I5POjFJRWjlURGaRdQdnzkQoTHK1TicCKV0DUH21bh6AFcQ3JOIlnbtMj3S4sbFhSYhQgGXPCK0 wnGhfftWYJCxQVOorkgPzCm23VZk0CCRV1+1zxHaShQGhyENzeIJy7To6WXLRO66W+TyK2O+ZOZM AlPf8eY1+8ouuyTpNqQ8U58+aRhg/vP111+Hm4UEBqXJli5NWK3AueOTjgnFgKFUjgISobkzxY+F WkIzB230SKmNI/KdCKVW4iJctX4ykMyiRKGCyUnz58/3Dp0Ujh1NNssL2B1o5PT007ZjJg6JESOs MKWiDoUr0A0kmaNjsJzifk6bQTcTReohR2uEUmyeFq4sV199tXzxxRdp2e5DDz0Us2YtHab87TAD gUQxQnTiiB3a3rrGGOqOzyIqQrOIitDMgXkEX56/N3wElFZx5VU0Qz4gUhShroc8F4hyLVfjgWWc iYZ6HvKCDh04JkWo+EMnUQ5Fmj11757hD06QlORvlpC0CKWLDvECdeumY4RpxcXEU6KLep4kqlG2 rrIQNnPAAQekYYQZYvFiO8uJg79lp4rQLFJAiUm5L0LVHZ9ZtHNS7uPPXE8C3HjbYMGQFOuF4q/F VZZrNZWShJJL6RAH+QQxnkRPkB0/bJjIxhvjCs/wvD2BCP3www/D1rFCKVKPwKKdK80Mtt12W3Nc /frrrzJmzBi56667pF9JvMny5ctl4MCBpkbo5MmTzXOExrj7MNP7g0h2o4LF559/Hn7+nnvuMe+L 7Cf/qffn0nK2f//+YQvzyy+/LNOnT5ezzz5b7r77bvPc2LFjTZcvxpQWUsiMBy3RlEXUEppF1BKa WVIQoeWsasQoYr2gwLSSOVAWiMo4FutIcqleaLLQ856uRXSYwe35fZKdohwTJ06UG0jDzgHuuOMO U1s0W+Buv/RSkRUrrFc7Y0YSJkKcL+KUq3ITn6233tosSZGjrvhImOhQ+owi84hQas4S9kJrTorO H3TQQbLPPvsYcXjzzTcbYYk4o0e8q7BAowQqWNBR6o2SxFD2FYrYDxgwQF566SUTEwps86ijjjL1 ammywC2NQzgXI1ixptIedMaMGUbcIkB3TiAckyaFpKSqVat6p6mNE29TRWh6UBGaRVSEZpYUOidh 4SiHWkOzQwXjQoPunJQss2fPNnVod9llF1M+iVCC3b19M5mC8umEXudYvCr7PvqCt04QN5kJatXC IpXBD8BFW7++/aAYFFI8qAOx2L59e2nZsqUMGjTIG6odK8KLuE7WYd3kf8dt36pVKxkyZIixarJP b7rppuZY/Oabb0yMKWLVz+OPPy4XXnihqak6bty4cAjUs88+a8qsIUYpI4aH47XXXjPrevToYTot 8VlYaj/yRONi7/854ogj0vOlU8yMjyy6Xw4mMBzPWk2l8qgIzSLqjs8siFAsnP/9F/MlzhKK6+Xr SKunitDsUMG4UC54KVkUA/o/sRTddNNN0rt3b2NJGjFihEnOcK5GLDy4NbGSugQ5XJy8HmsUdVHh b+9cQW/4vn37el+j9HtEc2mC34XJhZ73UgmC19IrPtLtuXLlSiMueA8WLRK/It/HcueddxrB4cZE P/oTTzzRWHh/LelSxraxXkUbb86SQpH6lERojrvjsTZi1eQ/vuCCC8LPV69ePVze68cffzSNEBy1 PKGOcIQTTjhBHn74YXnssceMRTOy7NfPP/8stSl4WoJbz77F/oP1lYXt7LvvvmZdTWr7lsDxPmnS JGN9PfDAA817KgVWW6r5p7NbEqEyCNBEYlVJDIY5FaFZQi2hmQUXCrNugspigHsHdwuUc8mrCM0O KYpQLDauG1BK1lAqJlCCp0TUZYP/vAkQlsTDDjuszPMkbSwsCUHg4o34Q8hxscZ9ibA75ZRTTIye s9Ljkq9Xr560bdtWOnXqZLYdy6UZ6cKsUaOGtGvXzlxQeV39+vXLuT0ZJxf5YcOGGYvUfffdV+59 LFhwXSgE1iri/4gn/O2338znA9vm+0SON6dJIEKZ8Hxaci6J1SkpKjluCUVs8j/FqxmLx4iJhROA zz33nJlQAVbLadOmmYx4JlKR7LTTTvJqSc0t9gs3cWSbTLBOPfVUOf300+W0006L2oCC/Qpx+tRT TxkxTGxqpSCuA8HoE7rR0G5JAcD+VSBJSZD7dUL5seP0N1fSgHPJx7BE0D+eizQXepKTcD2FQRx5 F1klwxASwcWdhI8k2r2QxYtowrqGGEKEJQWTDSzfntiK10krnZDEgkVxQ4pc+uA7UJvWce211xoB CFOnTjXWSdyazrXJ98SiiLAD3KZczGfNmhV2aYJzaW6//fblXJhY/XGxOmsTOLcn4PpkO1jEGjVq ZCzNTNAi3/dsSUtbLGFYUkmYYlIwfPhwI6IRDcBYI8eblFUpKJgIxTne3YSH78pEKCmwurFv56gI 9bd89YP72e+CZt9kX2zuTeSwiG7iiTiXkISAxS2/evXqcHiT//0kM7EPsq+SWEjMKesQsYcccoh5 T8OGDY11/Yknnij32Y888ohJkMI6yn7J+ytFEq540G5JAVBAmfGQ+yJU3fGZx4lQ72IbCy6yToSW ASvqRhtZ65l3UVcyBCcdxOH8+UmLQ9yhiNAKx4VmSYRiScR1uczbh5r4hAjucP/F1O9+RKxxsY3E WewBEevc5c6lCc6liRjFhYk7HDc5VtFo+D+X9p+33367dOjQwWyvSQLhxGeThOKs0txiVSPmNdZ4 cxYmBMSEcr6IgdvXmADxfZJi5UobY0pbpRykS5cu0rFjx3LPIxaZxPi55ZZbzL7G/8j/7oeJib/9 K5Mo9xiByf7O5ITjYe3atWE3/6hRo8wEDJe9E3v+9wKTL0JE+NxaceJ1kyZFEaqZ8VmkgOJBQd3x SuUy5EFd8tmhgnGh87z3pNT7PID/E2sjQtC5o4n7xHXZJ0YHnaZNmxrXJ3BxXoH7MAaxXJrRXJh0 BPvR9cOMAqL1yiuvNDGsO/o8B7HeV7ek7qU7bpgUYCFLynKUa1C6C2HiuqZFwYnQlFzxOR4PCrHa uvonEQ48R5ECFLCm+reDJTPSwlqjRIhHfh6TF/8+E+29TG7SIkAhiRqhkJI7nhADX8ysUkFUhGYZ FaGZJ4UM+f+zdx5gUhTpHy6UKCJigEMUERQVIyAiKga80zNgDqinZwRP9G9WzHqGM3CKp6KeCTNi DgfmgKCIYiKIIlEBAZWgIiuh//XWbo21vd0z3ZPD9z7PPDs709Pd093T9asvfqUHjDpB7yJC80Oa GfKINL/FJikFOJ9YeihyTnwnEx4yjBF87Wj7o+q6RIkPxQXJd0RI4u72uyjt/65LEwsmWevEaNrP E4tqXZgIS/YDVzIJTP51HnPMMUbI4jolXtXuk/9zvG4/S+mdQw891LhrsVg99NBDtfbPv79FS4pO SUwgxtZcn+WUGV+RiDu+eCkzEYqFpGD069cP84zXp0+f8IXuv9/zTjwxfztVqbRq5Xnffhv6th60 zbni8dlnn9V+8/XXPW+33XK8g4L31Vee165drI/owcGcswceeCD6h1au9Lw11/S8hQtj7mDmaCHn adEc+HoQP/74o97dlea5FkGJ57B8+fJay+rJk7dgwYJar/3666/eokWLar3GerSoDVwnLF261Fu2 bJl57r7nfo7X+d9loe94ptrfouO44zzvvvtC3x4/fnziHjFt2rTo6z37bM8bODALOyhkjRYtPG/+ /JSLtdLjBuf7iSeeSL3O887zvJtvzsLOVTjffON57dvXeun7779P/Pa+/PLLAu1YehS/JVRiQvND Cmso8U92thuYIU+sYrFn9pY6tFglucbpUpKKtOqFYt3r2rUg1m3ckEFFr8PcodRftNZIv4vS7yr1 uzQhyIXJelrU1DIMcnuSQNWoxiXtvud+zlpCXdwSPFH2t+hIkRlvrzFqXBLjGBmxhBYX8+dXl1FK Eeep9UMii18soXmkzBKTil+Eijs+P8RwyddJTmIQp3l1HJevEB8GBkRAGnGhpVK0XihSKDJOYtdW W4UuklaReiiBmNCKIqIrnqQ7QlBARGgeKTN3vIhQoZpMesiDiJb8kGZcKIk+tixQJOR8Ci60cuUe EWKRhrSK1HNvR9xKZY3iIWY8KEhiUh4REZpnxB2fH0SElgYxRShJOLh8KfnycZye8AXsIS8UISlc 8W7yW6zM+KlTldp44+r6tEJxEFGELnDCgsQSmkdEhOYZsYTmB2ou0lYtSXkaK0IpmGxbEiYQ0ZIf YopQ4iu3rBlQYrnksUzRNcl/noXKJEVmPI0CiBEkdpfuUpGReNDiI6YllLJSkWrCigjNDiJC84yI 0PxAvOH22ye1htLxwyaI1LGG0h2F1olyrnILdSepQYgFKSJpx4XGFLylzkknKVVTPUnwEzEpiW5B QTUyQ5F40OIjYo3QWOWZiB0lHMiXnCekgSQm5Rlxx+ePFC55Zru2QHcdEUorSbrbBLnqheySZlyo 7WUemQoLsfjHP5S6/HKZR9UBazgCIqBDlSWteFAQS2hxQYtsvGFt26Zc1LrjI3VLIh50nXWqjR1C ZoglNM+IJTR/SOek0iBNEUrf6pkzZ0bfToWFWHD54oV8+OFC70mRkcIVjxveTnBixYOCiNDiAlf8 FltEEotSqL5AiAjNMyJC80cmZZpARGh+iClCaXFp2wHGcslXmDseLr1UqRtuUGrlykLvSRGRwhX/ tRaSdKwCsYSWOBHjQSGWCJXM+OwhIjTPiDs+f1ADcNo0pX77LXQRawklE9bWiEsgIjQ/YKGkj7f/ +IdAdny3GhERS4S2bl3dJ3zGjDR2sjTZbbfqsNthwwq9J0VExHhQGgd0jCMoqT2K+7dNm0z3UMgW aYjQSO54sYRmDxGheUYsofmDMikEpCeJ67QilP7xdfqR89433yi1dGku91LAqklHGhLBIpJW5ySo wIkFcaHXXoubudB7UiSkcMdLPGgZEUOE2phQccfnGRGheUZEaH5J4ZLfcMMNE60J67jkOVfbbKPU J5/kcg8FSDMu9FN9bn+P41mosLhQ2Hvv6jy7558v9J4UAXhGOBiYh0NIu1OSiNDiI1fueBGh2UOy 4/OMuOPzi8SFlgYxRagt01RVVaU+w5UflQo9n5ddVm0NrXhSuOJ//fVXNWHCBPNcRGiJwzhL4uJm m0VaXNzxBUIsoXlGLKH5RTLkS4OYIrRVq1aqXbt25nkslzznE0tohfmmDzyw+l7/yiuF3pMCk0KE fqTfpxsXccc7JnHZByI1QosLzgdluCJ0r1qux2S6ZIEkJuUZEaF5RkRofsHKOXFi0qQXsYQWAUwE KFhPV6OIpBUXSm0/LB0MUBUEFWqwhl5zTaH3pMCkiAe1rni6cjVv3jzeusUSWlyk4YoHccfnGRGh eUZEaH6h28lGGyk1eXLoItYSOnfu3Fo3IwM15r7/XqlFi3K5lwLWCjpcxYjXTLtofQXGhcIRR5B8 odQ77xR6TwrEqlXV8d1J2nCmnZQEIkKLizRFqLjj84yI0DwjMaH5J4VLnrqTuN+gjkue17t2FWto PkgzOWnatGmJzNZIVKh1m0v5kksqODaU9o2UT0pi4bQiNHaRej2BNQlPca2nQu5IIzOeccAmqiZF RGj2kMSkPCMiNP+kEKFNmjRJ1AMUl3wBiSlCt99+e/1zqp5Bx44LrdDz+be/KTVlCtbjQu9JAUjh imcyM3/+fPM8tiVU4kGLjzQsodSGtQaJFB8QEZotxBKaZwjO4iKXFib5I0ZykojQAhJThDZq1Eif 2s7meSwRimUbi3fE4vjlBFEPF19cobGhEYvUr7XWWiYmNBbiii8uCL3gnBBOFYFY5Zm4bxC7Llbv 7CAitACINTS/dOmSstaniNAiYNNNqzvOzJsX+SNpxYVSHH/jjasT1iqQE0+s/jkEFYMoayKKUMp/ RbKGuYgILS6mT6eERnWIRASsO17iQQuAiNACIMlJ+YVSGmutVX1jCsGK0EmTJpkSLbWgmw+tP0lQ EnJLmnGhY/VnVmH9iEoFTyzoXHrBBRUWG1pVVR0TWmM5DyLtIvUgIrS4iOGKBylUX0BEhBYAEaH5 J4VL3pZpWrZsmfoqqHxPBYuWvJJm0fqff/7ZTCAiU+Hns18/pd59N2nRiPKChga4Zhs3Dnyb371N SkxLhEpMaHHBvYCWzREREVpAJDGpAIg7Pv+kEKFt27ZN1AUUl3wBiSlCN9lkE9WyZUvzXJKTooOX 8pxzlLruukLvSZ5I4YofN26cKVgOdmITGWIE6cxDYXShOIhpCRV3fAERS2gBEEto/pHOSaVBTBEK acWFUpOUgQo3bYXSv79SI0ZUt1N3iRPVUDKkyIy3ExiqZKwbtxMOYT6UfiqjgbTkEXd86SAitACI CM0/mfaQT0McCWmAVROLNHWEIpJW5yQCIxmk4vSdLzMIkz79dKVuuOGP144+ulpPkR9WVqSwhEo8 aJmRS3e8tOzMLiJCC4C44/MP2dAkF9XUAQwiaYb8n/5EQdGkyU1ClkgzLpSY0J/jqCexbquzz1bq qaeUmj27+v/33qvOv6tpo10ecE3MmkVXitBFMuqUJPGgxcWcOdX36ihF52uwIlTc8QVARGgBEEto YUhhDbUi9LvvvlM//fRT3QVEtOSHmCK0mz4vlNQhO35sHGu1nE+1zjpKnXqqUjfeWP2/7V5YVoYe WrTy21999cC3+b3PrlHhsTslgVhCi4uYrngmrlU1YTniji8AIkILgIjQwpBChG699daJ+oCSnFRA YorQZs2amdarECsutEJ7yFNnmxyaXr2UmjpVqfPOU+qRR6qN/IzFJC1hSCobUsSDWld806ZN1Tbb bBN//SJCi4s040FBRGgBkOz4AiDu+MKQQoQyCHWoyXAVEVpA6GjE8Y8xUUsrLlRPOtSMGUr98kvM HSx9OLRvv00ctFJPP63Uscf+YQ0tu/E1YpH6HfSkZPUQa2lSRIQWFxmIUHHHFwCxhBYAsYQWhmxk yI8bV6bpw0WEngwYU13QRCAEGxcayxJKD0tUGOe0guDwko+F8MQqesYZ1R2UHn20+v2yG18jitC0 XPEcQEJ3Ntoo3b0Tsk3MpCRbnqlx48bGEJESSUzKLiJCC4CI0MJAsWpiv5IkryTNkF977eoEpYqp 8F1A0uycNH/+fDXNX3MoGRVq3SYWFNH57LPVBQnwSKOnoKxEKImIS5ZUt4QN4Hc9AH5S09I37cz4 zTZTql69TPZSyCa5LM9U/YEy+5EUGBGhBUBEaGEg3pOYryRNs60ldOLEiXXbd0KFipa8E1OEbqkH nbWoOaQkLjQOhxzCta7UkUf+8RpzrbIhRTzoZ599ZrolgZRnKgMWLaqeTVFnLCKxMuMRTFwvNfca IQuUmQitX+gdiITEhOYERKPtH+55XuAyq2uR6WkRuSpkYNoCa6mimtNvRoja/y2rde6s6o0Zo1ZS UFHIGfX0ca4/aJBaHuN3QkzfW2+9pUaPHq0OO+ywaB/abjvV4MorY22n3GA8JTHp4INXU2eeuboe jFfp21PABKwEWR1Xe5cuamXI+R01apT5265dO9WiRQtjGY3DapMmqXodOoSuX8gv9T7/XK2u79kr YpyP76lJpvAOrJP6/M+Zoxqsu27S+0W9Gqs4Sa5pxRhXGmWWmFQaIlQsoVmDVns8EJ9hwtOlwdZb q9W1CK0K6ZTTunVrteaaa6pffvlFffrpp6YtpMvqWrQ0evLJ0M8LWWLTTdWaM2eqKqwUzZpF+kgX LTYQocT4RT4/Wnw0WLBAVc2dW+2jrmAOOIDY2npq0qTV9PErDxHa5MMP1e8nn6xWhlwP77//vvnL BCad33TjyZPV8l691Aq5HxQFDSZMMJbpOOdy3rx55i+TkFSfW03fJ1bX94mo60eQIkYb6DG/QRkJ rawiltACICI0Y5ix8vALT9caGsSqTp1U03vvTfSJDqKTXoZ6k8SF9u7du9Z7y/V7TSZOVMuXLi2r 2VsxskJPGLBar+jZM9LyiFDgvDGJaERXpCjboYWn3g5iotLBFb/zzuVze2o6bpyquu025YV8IVtX lmsn2T0hjCZTpqjfTzpJrSyXA1bi1J80SS3XE9g459K64xGhqT5Xf/58tSrCctYKyvjEmMQD4dpQ i62GZSS4soKI0AIg7vi0QWASw2XjNfmR4zrntSBR6qfen/6kOnz9tfqBrhohFz5lmhicsIS65Tss Tdu2VT+PHq2qKPEj5Ix6WvCveO89tTBikoG1WjNAvKc/tz3iMgqbb65W6fP5U008sFAe1P/2W9VM TxQXUAUh4HfMb/tbvQxsttlmgb/1VDTX95IF66yjVqXxWSH7bDBhglrcp4/6Ncb5sO74Jk2apLwG 1pwxQ9Vbc81I1wpWUAQn6yXzHhCiK1asMP/bmtQVT5mJ0NI4q2IJTQuE59KlSxMC9NdffzXZ0IsX LzY/7ijueE9f7Ms33lg1+uab0GU2r2nBNzkkC36ZFp+NcfsIOWWZFoWNY5RpIqZro5pSOZ/F6Am/ bJttVOPx42Pvn1DccO0sSzKx+LSmXBtCYcsY2dSW1SnVo+/lq5o3T3sfhezSaNo09XtNreeoLFy4 0PxtEaHN5+p62ZUR24EyHjEuLVq0yIxTv9TUI/aPYxWPiNACICI0NvxgsXjyw8Ya+qMeAJYsWZLU 9R5GVadOqhG15EKwyUhz58412/AjoiU/xBWhYEtsBdZ5DduOnM+yhHNalUSE2muEblvpxOs1nD5d /e6LGRcKR71ly9TqWuwtj1mzNVci1IVxivagjFs2f4HxTISoqtZCIkLzjLjjY8GP1gpQXBm4QuJm sbos23LLpCK0o1Ny5WtKsPg/L6IlLzCYrFZVZQaWqFgXfBxL6IoNN1T19PUUZztC8cNvdFmSNpz2 GokctuGjASK0Xbu0Pitkn4bTpqnlnI8YGemMLXjSYJ0IiYnpilAL4xbjF+OYFaJRPHhlCwK0fmlE UUalNESoWEJjQbyntYD+9NNPGc8esYQ2pqBxCHTN2FALEwhyyVdtsYVqQGyQvoEIuSWuNdRaQufM mZPohBJpOzKxKC/0vaLxxImhIpR7yISakJp0RWhDfQ9YLpbQoqHh1KmxXfG4yq0IzKUl1IVrj3GM 7VohWrGUmSseRISWGSSZWNGZDQEKxh2PCE0yA7Uu+cC4UH3+fu/YMak1VcgOcUUosX02+1Rc8pVL w2++UStatVKrQsp7ffXVV4nBPxMRKpbQ4iEdEWpd8RBJhGrRujIL3RysELXP06nMUBaICC0Q4o6P jHW7E8idrR8qA9PKddc11swwrAhlsAqiUkXLzJkz1bXXXpvIKIXXX3890frwqaeeMkX+LVP1wPDv f/9bXXDBBeqxxx6LfQ7jilBi+zrV9I0WEVq5RHXFt2zZUv2JVrxp0IAkGLGEFg2I0Kr27WN9JrYI zYIl1MLYxrhmn1ckIkILhFhCI0HcjA3i/jlJv/d0SBUXajPkp0yZEpj8VKmiBcvws88+q6677rrE a+PGjUvEzlL8e/r06eY52ccnnHCCyVg//PDDTTmckSNHxtreb/Y4x4ibSicuNJ0kqFSCnGM1dOjQ xHtY3h544AF14YUXGmFO4puQG6KK0M6dO6e3AX1PaKCvZyptCMVBOpZQa41s1qyZqh8hNjGbIhRs xjxjDONdxSEitECICI2EtZpR5iKdLPhkpIoLddt3zpo1q877lSpCYc8991STtIC3YiuMm2++WZ13 3nnqyCOPVN27d1cDBgxQe+21V6xtraIG49prmySQqNi4UGL+ooZvrFxvPbWqcWNTWzIqqQQ5IpV6 pYCl46ijjjKxqrQUJeb4oYceirwtIR6pyjNlnJQ0e7Zauf76yovYEEHIMfp33lDfp5enaQmNYgWF bItQW8QeKlKEllnLTigNESru+EhYAUFiUrZJVaYJ690aa6xhnge55Jlx0z1jtSxbaEsBOoGce+65 6sYbbwxdhkkDnYt22223jLcX10pphQWuLizZUaGcT9yJRVRBPmzYMOP2veyyy4wgR5AiyoUcoO+t DfV5r6oJy/CD8GCCAHbCEhfJjC8uGujzuVz/vryYVjUrQqNkxlNBox5Z7U2bprWPYdjxrSJFqFhC C4RYQlPi9oLPRbxMKnc83S7oogKByUmrraaWbbVVxVpD99tvPzNJeOWVVwLftzG8a665ZsbbMi75 GCJ0gw02UOutt555nuu40CiCHAhNyIYgF1JDP/ff27cPtVJ+UXMt4X7dOs2uZ1IjtLhohCt+001j f8664/MdD+pixzebLV9RlKEILY2C7ABo2wAAIABJREFUUyJCU2J/jLb3brZZ2aoVSlPVnzfPZNEG gUseEZMqOWkpzbYrDEQ6lrxLLrlE7b777nXeR3yutdZaatq0aYnQhnTBEtps+PBYn8Ea+sYbbxi3 K1bHSNvR53Odu+6KvX8I8gcffDBUkAMWl2YhmdpCdonaKYlKCo3SdKebzPiYrl8hOpwjQlui0mvM GNW4qkoNv+++WNv5+OOPzV/Kud2X4rOt589XffR4lGq5ZPTs2TORb2BxLaAYX1aPUee05BERWiBE hKbEFaF+mL3ef//9RghlwllrrKHeuvZaNT4kueC7774zf7/88ks1cODAOu/vMHOm6vb22+quLLnk cdcee+yxGX+vfLHDDjsYazHiq3///nXeP/jgg9Wtt96q/vOf/5jBnsQkWq26opSEMxsTBQhXW2LJ vreaPi5tvv5aLdefbVDjCkv1uU033dSIUAYyLA3+dQZ97od27VTrCRPUD9QX1ecgaF/cz1lSCXJo 06aNEeRCMAy+VDXIhiWISFsice8bNizpcuP1BNIvCKLyun5wR3hV3z9S0aVLF/XEE0+ktZ1KpU+f PsbDEPVeuP4PP6gpbdvGzh3461//qnr16mU+l+qzTfT955fGjdPOTyAe/MADDww0anDd813FElr6 lIYIlZjQyAT9KP/1r3+pTTbZJO3SKpYlHTqo7fQN5dcQiwbJI/vss09oSz8SZjromXT7LFlEhgwZ YsTTTjvtlJX15YLVVlvNPCxkeh9yyCEJixI3UjtwnH322erqq682cZPNmzdXTZo0qeO2RqS++uqr if8RrPvuu695TsLP8BoL6NsrVqjZ//uf6n7GGXXeC/rc//Sya+vzg3hEbASt0/+5s/R7V2mhef4B B6gp9euH7ov9XEPn5hlFkPP6QQcdZK4X4sDGjh0rLvoa8HZwXWUj9KbeVlupvz3+uLonzXjPSNvQ E5ZeegKqUrjkKVH2l7/8JWf7Uc6Q2Ojea5Lyzjuqq77X9OnePXc79OSTJgwryCARBYwab3PNBGBF aMVRholJpSFCxRKaEcRy4W6l/E9GIB6HDlW7ZZIgct99asCJJyoV4tKPw/CYLudCwICK5cCy8cYb G2ujvYEycFh3Ep2nbrrpJiMwqDIQFB961113hbqpa2WP9+2rtiYeq6ZQdLLM8oze0yLofYTn8cen /Bwu+GSC3BXsXbt2Veecc476+9//bo4DYqtfv36h+1KpRBYdYeCVmDVL1cMdn+m6wqDI/YIFqh4C NMU2KlJYFAIqnWy5ZW638cMPStXEmgtZogwtoZKYJESHGoE18WFps+OOSo0dm539KRH8MUvuQBsU z8RrQQKU+poUsY9Evo5zt25KffRRysUIK8Da6pZnsoIcIQoI9kGDBiXeJzaVkk1kyhMqQOkqIcsQ 48fvOpdxdVRcoB5lrkSuEA9K6DE5dUJkcsKPPyq17rq53UalISK0QIg7vjggm5Ibi9M1IzYRRYtQ Fwo1h7mn6pAvEbrDDtVCJgVYMilMn0yQQ5Aox/IrFrIcwW+R32QuoQ5sx4653YYQnXxYQUEsodlH RGiBEEtocYAQIG4sRmedOkQULUKGbLWVUhSSX7Ikt9vp2pW6TqSs5nY7Qm5gosKEJZeICC0uRISW LiJCC4SI0OIhU5d8BbrjCwIWxS5dcm91JvueWOEKrf9a8uTDEkp2c5pZ9UIOoN5zSGOCrCIiNPuU YWJSaYhQcccXD5mK0NatlSIRZcaMrO1SpUB1A3qoR6aI4kKJcb377rtzvy9CdObPr7aUp1G0PBZi CS0uHEvoEn3+bQH6rJNEhOZ0u+VMGVpCJTteCIUyGNSqbNu27R8vIkLjCKEgrGgJaeMXuF3BCLmw upqBIELzUW/Rhlj07Ru6CGW79tprL7Vo0aLc74+QFAQABb/XyYcrHsaMMZby3/TklcoPybqCNZs3 Tx35889qraeeyv1+lREn6ccyPcmr0ue1eZMmdaomzNPHleQ+6PPBB2pOs2ZqI1VdsYKER8r4pctp p52mzjzzTLUVIUAuSURoqu1SW/S1114ztYapLvKDXhevtWrVSvXo0UMdccQRlVnKS0RogRARmnMu vvhiMzjR/52CxAiGuXPnmhvAN99888eC3GgoIk7ZFX2zSwVFp5nx1qoFaUWovpEEEbjdMscK7w8/ /NDccEnQoZXmLrvsYm68aYHAOOec7O5oEPp8/j54sFpND4CUAxOKA34/ttpAy5YtTbkrKhIkBACD WRqueOo3nn/++Wro0KGpF6ZQ+e23m9/7tIkTTekxasSG0XDhQtVl2TK18r33VOPGjeV6SsIKfWxf 0ffK5/T94rVGjdQjZ52l1tTHjPa/O7Vpow7U57zf5purhvpe8tuMGWrtSZNUFzqjHXigOn6//UxT kWyAIKRGdB3SzI6nOP3JJ59sBOhVV11lCta3bt3a3CMZG5577jlTZeN2fV3RhMW2i64IRIQWCHHH pwWz33feeUdN0jcfuk9Q7Bvr4qGHHqo6+txj1KckexnBeMopp5j6lTsHtddkQkAHH2IAI1hRGHAo NF4LBr4MZt7lBufolltu0ffsH9UBBxxgugVxA6Zn9w033GAK8lMzc6ONNjLLhbm13333XSM8uIEb 2rVTq/R6bjzjDHXxHXeEWywyRQ9snh7Qxmuh0blHj8BFKNF0+umnq+uvv978P3v2bLO/CG6uj/XX X1917tzZ7FtYs4OigGYQCCv+UjC75m/iUfOeyeV3luN5vYifT7mcfn01LTS20X+nPfOMWrxokfqT nqi01MdwdZIHa5b79s031Sot5k7R18NkfQ84S5+nhx96SG349deqKa5QfU9Q554b+xCsq4UFgjYS WORqGia8cdttRvzu4Pvtc60//vjjRtRSkou6sS30PnLP6tChg2necNJJJyW1oFYaTAQo14YwO03/ veeww8x5AX5rNLTgeA7Sx/O///2v8UA89thjav9nn1Wt9TIb63v3GCzUqrpyBU0xvtbHnHtEN31/ ZvKCMcJ2a+P9448/3jQToA0nNY0RiPxuadXMcjTYoNHFYCak+pzerielq/S5/K/+f9SoUWb/zjvv PH1bahf6vSZPnmwaUnDfO/HEE+tUxaDpCp2huB8SmsQ4NmvWLHNvLBn4XdNam7/2EfX/xYtFhBYE sYTGggH+zjvvVG/qQWj//fdXffv2NRYQZpG0QsTKiLBBELhCE1cvFrjp06ebmo72PZ4/9dRTxip3 6qmnqu/1rHSmFkLda0Qo4pWBBAHF7PT99983NzJuFgwkWFi32WYb042HG8eciRPVnR98oBroG+g7 WogwAPF56kCyv5UCrslLL71UTeR46POF9dl/08Wq8fTTTxvL0x577JHo3RwE1i5qa9KUwJY6+m6D DVRTvX4ItVgoZa6HD/Q5YVCJjb4pTtWD0qZYu/U++AUZj4b6/J7z4Ydquh6krtQC6Bv9/6EHHaRa t29vWnrO0uL5qgsvVIu1EL+qWTPVhxagrDuOuEsh3KKKwKDt1HM6kXmcI/vQIsu84/xvn/uXM6+x bNgySZaz2+Hxy9KlarH+TT2sn1cde6xqWr+++kELiXl6f5tpIbCBnsQ0WWMNte3Chaq1FvlbDBum uujPb6FFwXJ9ve2oRf8yWqp++636Pz14H6evkTjCAwslE1rbHcsKD67n//u//zOfR/hEER5Y47hm uXYRQNxrXOvnyJEjjTeFbdHm1W6zksG4cPTRR6t//vOf5t7uv2dwnjA08GAM+POf/2zuMxbux3hd +M1PmTLFWMZv0xMEDBR0KEPgImSZ7HJNzJ8/33Q8o6PbjvqezzoRtda48IyeCHFvYvLI9YOVsiMh AXq9iwcOVFt99pk6YvPN1UR9nbyi3z9N73MXPfH+lUYJTEhqRNZifb0+pbc3St+nOlJblqYoIaKs nv7/fP33ED1mvae/x0F67GmKZy6uqMv3/9xHOF/cn/l920fU//XYXG4GHBGhZcZHH31k+qkzi3z0 0UfNgOEHIcgggZuDG1Dv3r1rvT9hwoREj+iZM2ea1orc7LB+8PoWXbuqKbfcorrpgYnYI+J6Lr/8 cjOAIEjvuece9VlNGSf+Yv1gG+zXBnrA66dvLg20sNKqV7XTN7iB662n6C48Vt80q/bcU7XQzwfr z6jDD/9jILY/Ruf5gK++UpvpfWj58st/DOY17yUGcTuouwKB9/3LF+DzB956q2quBQOW6qD6mMCN nQGH40e9zK2JsSKWz39c9GNL/f8uLVqosQ8/rHowgdCvva9v9L1rRIW1WHDO7r33XrM+3P9YHxlo CMlACDOQIIovu+wy87lnn33WhAQwCcFqGzRpOKdTJ3WnPr/NsLAHCLJlelA7RYsNBsCBekDjs0Hf GUHeT090HtYD3BB9TZllooq7NERgXBFZSJhEnnHGGWoNfQ1wrpj0uYKNCR8TFsQJ4pHzhfWL88dx 31//dt/VgmHEiBGmPextXCf683GFBxOjRx55RF155ZVmOSs8mOhiyVy7pksX+8E9AXHJPcTPjBkz VM+ePY2ADWtEgFWMB92ysJTjzkdcVSp4qjgeD+tzd9xxx6Vcnt8zHhUmGTbUCm8JVlSsiFwLeE7s ungdjwz/0w6ZsYJrivND212s1FhZsUwjdl2I4Ux05mO8vuYa1UKLpp56YrTw66/VhvramcF9XV9r a/zyi1pFSBeJcfq3tUr/tp588km1hxbMHeneFVGUddD7OE2PKbd98okaoCcpq/F7iCPqCvG/UIvS EKGYn4lDvP/+Qu9J0cLM8NcFC9STWjiM+Mc/VFfEgB6AwuirfxRHaDF4oR6Amusbzkl6hna7voEw 0Oyw7rqqb/fu6rcnnlDnrLWWukm/TkbltZtsolZq8dqySxfVQ8/qsFIgarCuYkVFAGPZQFT5LRYM XtzwuMFh+dLqR+mNqXZ6PfP04Pqj3vcX9ey8w957qxb6pvW8Fq979+nzx+zRziid55+//75ae7PN 1Fq0A3Teq+dataxly7VyudawmuU5fmGfT1jFknw+2faDPn+Nvjl/MnWqWjF5cqgAdcEVuUAfo276 Bv2wFqXH05LT/Y416/6vHqSWEweqReNKPRAcqgf6+jUJQ9ZigeXkiiuuUH/729+0zj/cuLIIx0CM bL311uYaQGhaEYpFi7grRA2TCtzqCBFEEKKR2NVf9M1/ya67VpeFCuARLTQW6fNKvNdaSTq1IIhH 6fPKgLn5EUeoTz/9NNHWs5JZuHChEV/XXnut+of+fQfBBA9LJOcVCyQTBrwi2RYeiF1LLeGhwXUO iEbuA1xbQbHdWOJwrT7//PPmuksF22BCjEcFty/iqtLA9Y41GFd0FAFq4ZhxnpjIYvHkt+wmLrkT mcY1MaVYRTvpiSWTD9z6t+oJM/cN7gVcH9wnuJ90cX7vXEOETxgwHOn7B+eM88s1yLof1ve7E265 RY2qCc3YvcaqN0JPVu586SX1+euvxz4uvfS113/LLdUuehIUK3FTKApKQ4S2aVNtgtaDkxDM73pW +c4LL6iz9Ix3G1xtEY4VFsebDj5YjRg2TPXQN7gj9A+5SZMm1TelcePUyqVLVTdmtDXr6jBvnmrC zFUPTNOPOUYNffRRtd1226k+WixyU+uuheuQIUOMGw8ryOvODYVBlIExIUBqXPm4iBnkWM9wLcZO 1DeRJq1bq9c431hCkzDigQdUG738enrgLCWwaP1XH3fcalEEqAWx97wWENzQDx8/3gg1P799952x GM2ZMMH0b8cK8p+LL66zHOcBa6fdPgMEFi8snoRjhIFljIEFyxkuVOKx1ktRC5DBisQ04j+TCVCX pfraw42IOMbKXskQxoI4JB44TIC6rLPOOibOj98jgu0B/TuhvFe2hAfrstQSHjX4hQdZzS78T9w5 rv0oAtTC/jHx5bpA/LRgIlZBIOpxnw8fPjz2Z7Ge7qoniVikN49YsxXPF78/znHXrl3NNcK1Rb4A v2P2wxWhTCC556yoSVDEo8a1Q14AMZ6ffPKJseAGwetnnXVW7O8F3MOYfDH2iAgtPUpDhHLzJD5E COVmLfom6799Hn881ue4jS8aPFhddNFF6u9PPFFLFC2aM0dd/NFH6ogaC/QzWswwAG2lbxY7LFqk jtOiFZHDIAe0lcRa9uKLL6ptt922lphB3BAzhqV0S/05blBYbigbwk2OGxyhAZUArnBuuIi4uCDW GYQJeTgnIPOdmE8GA44rcby4XYNAwCYTwH7hYHEnDbhvPSdWMgjOK5Y2xMO6MTNlX375ZWMto7c8 AqhSQQxi+Qo632EQxoEAQLhhdeS3FoUowgOLqCUd4UEICHGmnN+4cN1hgSdOFIt8JcGxJ8wqKMQq CsTmEsbhij0mGW5Mqfv/YYcdZsI/GBsAyyXHnvheQimeqimjxeSGB/clfqdcE9wjuJYO1pNtrh0m JEwamtRUVHG3w70G6zz3tHTBS4DhQyg9SkOECknB6nHHHXcYd3g6YF3B2sQAwk3e4r9BcaOx/xP3 heuN2bkVCMT02FhDrCMEvtt18Fn2ce+99zZCCRGEUGJ2jnDFzcdrdvly7hVO3NyFF16Y9ue5sXOc w0TJMcccY+LsGCh2imglxu2KtZrJAmKRAQdrJCJz9OjRxoVHPFrcSQNufaxrxP7FhcH2mmuuMXGK XDuVCDGXiA9EW1z4PRGvzTF0BV+mwsNdPh3hgaufmNV0qyAQn44QrjQRykSOe2y6YCUkNIKwHpuA SLUMdyKJkcBayTmfTDyI5QdCK5hUcB/A+2GvAcI+7ISWCQeTU+739rxjvcVAwbowRPi3y32HbdpY 4nRgkrV48WIT9pWuSBcKg4jQMoAbA1Y1EgPSgZsJFi7c564IpTabW0uOwcwdvLgRubFJxJmRRINA wQ0I7s2GLFhiEEmSsK60l156ycQgsjwzYnszy1YNu2IDAcdNGktxuhAPR5UCjmuQWOcY457CAuli LRZBIt+6T3GTEgtKXCHXAoIUAcLyWMSCJg3uuv0QSxjH5eqH5DmqLHBN5bt0E9cjg6b71//c/s+D 85Fqeff9KMtTsYDJHMlF7rZ4UL0iWc1N4PfI8UME2BJH2RAe7u8zrvBAtCYTU1jFWD/fl+X9f1kf YSAIbK5D93173sL+5699hP2frfVEWW+c9TAJwOLHOfCvlzAZLJ3J4N6KWCMcyIpQ/73ADdPgPGBM cMcV1uGvoOH3qPgTlrgubBiOXdbdLhMdytIFwXd6PKJ3j+sCy3+cEKdyIcxzVQrU81L503IIWZOY 4LkYpShx+tibfNiPDxFBDbhkkElP8HpYzI4fZpwMULjjEauFAJFkB1ArrvxCK9n/QX9zvSwDMwNo srjLKCAWmPmHzfqDRBvXib1GVgQUlmd9CBW7DPvKgGLFrh00giYN7rpdsKAymG2aQWtIXPIIIERv vkRgAW+LkaHE2sUB8b5+9ttvP+Pt8FfBCAKL6cCBA411K1cwAcWyy6QmCCZAjyVJqhTqgicrijua LkNYubFgp4LfP+S6PisTc+tp8YNhw4Z7Canhnk4yYjqhXoWioMrPtmXkhr9cSjBlTNhsKMqxRUi6 Wa+pQASRGV3Ii91eN6V27YTV6swmQVZDVyQGTfqs9dpiBx+/1dS6zdz1hU2ASL5Ku+NTDYiVZPVR SxWOGZMT+zfoOYKf729jeO2D96PWdOW3zXmIAtUTeOQKBD5hHsmSiph0kPCSbFJHbVKuK65Fa4F3 3/f/74b5RP2/2NZDwifVEfid+tcf1bNCQmHU+OB8NQfgWmBSzLXud8lTZSFKe04E7IABA0zpwUqG pLNSEqBQUBHKRUOsWJ2OOkIsyGQlgxYXbRBNIrTXxJ0WZTkLN79CX+xY6XA18sOzVqwg9537Xqr/ 4y4bdTv2LwNwplYmbtZ891IoXcQAyQCT6ToIFcBl54o0K9R4cD1GEXX2b7rLB20/7vJx3IXEV+K6 psxOuiBisXJHIdfCg++O6MAFG+ZBIeyHRzIoDYfF1t/5rZwhlpaKB4Gd7CJA7gCu+HxMguPA74KS foScuElvQMWOHSN05iNmnHhmKrUIpUVBRSgXXzoJC0JtyH4laD3KjDEMN1i9VMDah0s+avJNscB5 Iowh3bhQSuhwYy6F5C0mKlxbmZTTYYJFOZ+K6hFdA8eP33cmIpTPdwmp31oIOI8kNKYbxkPBerw2 US165QKJRfRNT1eEEvPPdRC1TFo2wDWMxTpVuB2JatRA9YvQKGBdp7ECyZpC6SHl+8sA2nDSmSST ODZiwVhPPmF/iY+sNJh4MetPFwp877PPPlnco9xBIhPJSemC+xBXsr8WZaXQvn37RPexdLGdsvIF wiOZ9ZsuWy+88ELa60dMkdxUab3ksYrTPpVs8nQghpgYS8AqSgUDkkWJKw6KxwyC0ljUF44KDQ+Y cKeC+xlGhfvTaEhD4w0mNNxrhNJDRGgZQPYis8103bzc1N59911TPslCMgmzU25aUUs/IRjiuENw DeVb+BYDhE3QGhELYVwQFJRJIlaqkKQSGhbc6JkIblyQiPagzPtKgEoEmQg2stSxJGN1LxbhQRkn 2n7ikk8HrOKV6HYlEZSyeLRLjQuWQu7zh9c0ACG2lMkNPeUJaYiaKEk1BgrfA9fQXXfdFXtfgsBS StgJZeeoMx2VYcOGGXHNZysxK74cqMw7e5mBW5ayOunWniRbklJLNsYToUM5FTJqKfFCAlIUKOfD AAOUdbEt/ITaMGunBA438zjxkpS2on0hCQpB3ZKAwYRi1AzSWE2iQqWKqOcZolo4KP1F9iuTmriQ nUtP8kqrB+mCCOX4pWtNRrBw/Bigi0V4cP0zwfWXEIsCWeBVVVUVOXkFrH50PcJ1HRWW555AvVkb R044BJM7JicUwKcxAXV/bQIg1Qv+VdNSE26//XbTnpkwCEpFYfCgmgr5CHbcsWEzVDdg4mJheTqm cb6TxSbj7aBNLyXdqHOczLPHe1yHjFPsl9vFSygtRISWCQRlIwJtD+eoUCKLWBpuHpbJkycb62qP Hj1MKQ/WieBwO1qQVEYbRnj//fdNEWsE1dixY03iDEKJslDUq2R9QL03Bh9upCRCWQglYBuVFNPD cSG5CBeULbGVDAQZ8VxMFDiGYRDgj1BlYMBaGRXEok1YyKaFgwkSHX8Y6OKUpWKQoekBhdJxSVcq WIApHk8dSAb5OCAqiBW3rT6LSXhgkcVKGycGEKsXQpoMaJorVCJMPrnXcp44du591A+/IY4V9wsm H5Tqs+Dh4nfO8WRyC6zLusPxXjCB4T2qjzAZxNvGerjf0/WOEA+uISakwMQa6zbWTDeEhHPNMkym MJYkAyHKeEGJLuLeseDSBpaJBwnM3EO4n3AvZDyhOxfVFITSRURomcBgT/s8Hvx4owgbaoNyw8At 7iaO4JZnEOHGYYtSk7TETYksb95DZNoexohTXmdAws1GrBbr4DO4/shqZlv0n+bGicClxApwg2Gm ThcU2gUyaFYCnC9uorSktG02g2b+DAAM+lQAIDM2WVcRlkVIUA+Sm7R120URGli/Ebq5sHBQXB8L FmIySgww1y5xg9QWHSDtek2dT5LvsEbiUo+Cba/J+bS/7WISHohIOodxPVGAneXCYHKLyxUxjkW9 1BIosw1VTMaNG2fc6/xGzj//fDVmzBjz2yLEh98y92cs2PyOCbVym5AAk1WOP+cMqzgWSBpD2Hs6 9yNCaYi/pWMaY4pbhJ7rgvs6v2m2g0Dk3HAP5/pg+xaEJElVhCHRlS8V3OvYN7wz3HfwBlCWioQq fgNci1zHNDoRAVr6iAgtI5glk8SBy4XAfcRdUA1NYjwZ2ChuzE3I78pg0MKdzqBDJibiERcaM2k+ y40JocMNC+H01ltvmZuahfgerC3cNLiBYNFi9o7rhBqEDIY2o50bGUKI7bDOKDepcoL+2gzgiDms nFioGLQRdrTfpEg7Azo3d9vFJgysqggPrF1k0VpRG0VoPPPMM+bayZWFg+/C9Yk45nNBohXxiQBm +1gAScAqhQoA+YBBl980gz5dlMLgnN9yyy3Gs4AAdZsEFJvw4HrFa8KklHPOtY9XhHsPXhTuK0yE yKZnv4n7q2SruAsTfe4deI+oFsCkn3so1wghUUzuEe4cQ37TQbAsMdeIPZalbjflvLjH85umexrX Bo9UXc8QxHjigjLvbawm41LUSRS/f5KV+AzfAU8c1xGVQbhn8Z7EgJYH0qaozMDCgAuGAYi4IaxV 3LgZjLixYAlDDHCDwb0VlmHKzPPqq682IqNz585GQDJgMUh89dVXRsByY/rwww+NJY+bj+2wEQTb Dspwdkt3cMMptcLz2YDYWcQ5pXQ4noQ5cINFyBPPZ2N1Ob5MCJKBVZrBGlGIuEckcN4QfuAKDQaN MKHBgOQKDc4f5xixYUFodOvWzVxzqWLUuM6IUWUd1PTj2sKqQZkdJil8dwYbJju2frBQG6zJCAqS S5gQMjngL9cHooNrh8kEy/D779SpU511WOGBhQnhweTQFR6EypAIxbWRD+HB+7hXqffLxInrFC8J E2G+F6ElvJZJm9tyhQkakwkexFAi6KMm8BE2Zetv8vu1HhbOOb9lMs3xYBAGwm8Xa6sfPmcz9blv MInEc8b1h2DkWhOEVIgILUMYALiBELyPoMGKwkCBlQxxkapcC4MAohW3DzcabkJYLbhB4ZLF9c7g gCBhYMSd5ofBxS0lglBlYERcYAXjIa1a/4ABhXNju4hh5Xj77beNSOD8MQhzHrAupYIQCM4JAhFB wUSikEIDuH64Jhkob7vtNhN2QVgAgxWxqMSk4m5FhIr1MxwEKJMKjh99tZm8cE45hlwjTCzdiYJL MQsPMr+Jk8ZDQxy6tc4KuYEJJJMWftfch5m8AGWgmCDi9WByisDFk2Inwm77XqyRLI+7n8ktk0ss slyLGDp4zTZpAPe5IFhEBZQz9mieAAAgAElEQVQx3CyIn+KBq8YtwZQMRCgJIQgCYoxw5bIuhCQ3 JFzswDK424YMGZLYnr1BEddDLBfuV8q7IKRYHksqFj1iRBnEXMFh29AJ1aV1cEtC1G43nCuOLXFS HEsGl2ISGi5YWIlFBSZFxIAK0eD8MBnBaxCn6oAID8HC/ZfubVxDbrtejAtMKG3bXxtyZcEjZv/H oEDyo20XzeSXSRKhPjYOGcODnayy7ky7xQnlh4hQoQ7E73FDsVYwVxgiRuyggmWV0k72hoVotYlM WMfIxsXyZW9ICBhqE7r9nu3yQKs+EaHhEJdJvN/dd98d+D7CDqHPsUe4EgNou2ilEhquWMiF0GBb hBbgAhYKgwgPwSWszJu9DsD/u3YnJNC4ceNa79u2rO7//vcFwUVEqBAINwuSYoJed3FvWOC62LlZ +ds14hYOW14sJsnByomLPgxcmtT6xD1LnKVbxiaV0HCFQy6EBtsmDlVEaGER4SEIQjEhIlQQygys 10EkExp+oSBCQxAEQcg1YnoSBEEQBEEQ8o5YQgWhRKCea6oSTcUKSVBhsayCIGQOna3KKaaeLknU QBXKGxGhFQCxf8QTRun1XUrEaQNZDkQt0VSMuCWahMxBbFC14Oijjy70rmQVhAcJeEI8KOhPXd9y guu73L6TUBcRoRUA2c504KEIeTlBUhNZ3pUE1o5+/fol/ie7nVatQIkmt698sb03aNCgDL+9YOHa p9YrjSPKDerFCvF4+eWXC70LgpAWIkIrADry2N7hQmlDeR1aMloox0S3ISADvZjfS9ZRS4gP9VVT NZ4QBEEoZkSECkIJQdklGg+U8nuCIAiCAJIdLwiCIAiCIOQdEaGCIAiCIAhC3hERKgiCIAiCIOQd EaGCIAiCIAhC3pHEJEEoci699NLQnt+lyq+//lroXRAEQRAKjIhQQShC3N7u8+bNK+Ce5B7pOy8I glCZiDteEIqQvffeu9C7kBfo/POXv/yl0LshCIIgFIB6Hj0dhZJnxYoVps8urc7mz59f6N0RssAP P/yglixZUujdyClrr722WmeddQq9G4IglBCtWrUy3qImTZqY7mFC6SJnTxCKlPXWW888BEEQBKEc EXe8IAiCIAiCkHdEhAqCIAiCIAh5R0SoIAiCIAiCkHdEhAqCIAiCIAh5R0SoIAiCIAiCkHdEhAqC IKSAhgGUQcsGM2fOVD/++GNW1iUIglDKrH6VptA7IWTOqlWrzCBJ2VdpiSjkGoTUHXfcoTp27KjW XHNN89rrr7+uFi9erFq3bq1mzJih/vOf/6h33nlHvfvuu2rRokXq/fffN3X93LJTo0aNUpMmTVKb brpp5HUXglNOOUVttdVWqmXLloHvT548Wb3xxhtq6623Trqe22+/Xd1///2mBmyPHj1ysauCUPZw X6DRRYMGDWp1lxNKDzl7giDEBtH17LPPquuuuy7x2rhx49TXX39tnn/11Vfqyy+/VH/+85/NA3G2 dOlS9cQTT9RaD4IMYRpn3WH8+9//Vo8//nimXy0tEM7vvfdeyuUQqtdff70699xzM9peNr/rxIkT 1XHHHZeVdQmCIMRBitULgpAWe+65p/rss8/UJ598orp06VLnfToh7bLLLon/e/furfr06aOuuOIK 0y9+4cKFRljuvPPOsdZdVVWlhgwZoqZOnaq6du2qjjzySDV27FhjVV1jjTXUd999pw488EDzebYH AwcOVCeddJLZJ9aJe33fffdV3377rXrsscfMvuy6665mH+GBBx4wwvnJJ59Up512WmLbeBvuvPNO s/5NNtkk8Lh8+OGHZjnEHRZh9mHbbbdVDz/8sJo9e7a69957zec5BghzxOuWW26pdtttN/Xoo4+a /d9iiy3M/mLpwXXPZ7Ce8l3xdrjf9cILL6yzv1igDznkEGN1ZptvvvmmOv74483+vfTSS2abbdq0 MfvGutnPa665Rh1zzDHm2CxfvlztvvvuZvlBgwapfv36mW5sw4YNM1aozz//3EwSgs4FFipBEIQo iCVUEIS0QERh0bvxxhsD30dgTZ8+3TxoP7rxxhsb4TNmzBjzPi52rKQIrTjrHjBggHn/1FNPVa+9 9pqxrnbo0EG1b9/eCDHEJWKTcADCVBBqCK2RI0eazw8dOtS8zj5hAezcubP5+9BDD6mnn37aLMN6 sTYedNBBaq211jKvIf4uu+wyE+4SJkABS+4FF1xgXPcI7NNPP91sj+dNmzZVvXr1UptttplZju+y zTbbGBGKaO3WrZs655xz1Mcff6yef/55s76bbrpJtWvXTp199tnmWPm/a9D+jhgxQv3000/mPdr4 cqzhhRdeMIL1sMMOUxtssIE5NkwUWrRoofbbbz/TDnHChAlGZFqeeeYZ0xKY780x5f199tkn9FwI giBERURomSFWCCGfIFywkL3yyit13psyZYq6+eabzYN4UDjggAPU8OHDzXM+Yy2PUdf9+++/q7fe esuIMGsxHD16tLH4IaA22mgjI+oQgAje8ePHm21jWSU2FSGJCMbKx3qw3iGoEHRY+1599dXEts47 7zxjnWS9cNtttxkReMkll6Q8LojBgw8+2HxffpPExBL32rhxY7Ot9ddf3yy3//77G4slryFOt99+ e7Ns27Zt1axZs8wy6667rvrggw/MviOY/d81bH+D4FgiuLt3766OOOII8/0333xzY93kuY3BDYP3 r7zySrOdsHMhCLnGjnMy3pU+IkLLBBuczV8J1BbyBYMA1rBbbrnFCEaX7bbbTg0ePNg8/vrXv5rX EF24hhcsWGDiKBE+cdb9yy+/mL/ff/+9mjt3rhGaYfGMe+21l3FLI0LPOuss9emnnxoLH1ZIxBSW 0LXXXjuxfLNmzdSyZcsS/2O1dMEaSRIV4isV9ev/EenUqFEj494Owt0Glk8siuwv3w/rKSAusaL2 7dtX3XfffaHbdNfFsUO0+rECNxl81m7bDyIayyfEOReCkC0Y30SElg+iVsoE94cZ5N4UhFyxww47 GGEXZA31gwUQixlFORCmqQYR/7pxszdv3txY8ohf5IELG7juyaC3IEKx1E2bNs24u8luJ76R14H9 IJbUikriJLE0hoFlk/W4CVPZ5LnnnlNnnnmmiel03f0I46OOOspUDMCdDv7v6gf3urWkEj9qwRpL zCr8/PPPas6cOaphw4ZGkLufZYIAhAggXINIdi4EIVdwvQL3DjG4lD5yBssI+4PE8iIIucRvcSc5 hphBe+0ls8iTlEOWeJgrPtW6r732WtW/f39jNSSxhkQb6Nmzp4nrPOGEE8z/iFdiGTt16mT+xyWP K96K0B133NE82A9EHklSrDNoH+z/l19+ubGm+gW3u7xfWPO/a7kJs+LgusfVTTko9sWuD2swr2ER tZZG/3f17y9JSAj9ww8/3IhV+x4hB4RD4Irn2PEe7nQszYhsrMXE6RKvymfvvvtuteGGGyb227/P YedCEHKFvQ9Yi7xQ2tTzgnw2QkmCyw+rCaeUDFc5tUIuQbi4AwHXmytS/O+7cK0ms9inWjeQeIML 3V0P2dq4km3ZJ9bjegmCtosllGx2ss3Dtu/+b13VfpFtl2Ffedj3Wbd1z7vP/csBv1/+x9rDdux7 ZO/jbrdWIP93DTrWfFfeZ9D2v491kwQmu372A2solk0L7nbCFtzPuvuf6lwIQrbhd0y8N9ctoSFy vZU+IkLLCFuo3g4oUrReEARBKBeYiDF5QozyXGJCSx9xx5cRtoMEYMGQeBlBEAShHMAaj7UdGOdE gJYHolLKDFxvNj6MBAP5oQqCIAilDOOYHc8Y2yTvoXwQEVqG2B8o8WNujJcgCIIglBqUUrNePmJB hfJBYkLLFJItSFywz0lsCKv9JwiCIAjFhvXo2YQ8DCxucp5Q+ogILWNcIUqGK+VY7P+CIAiCUKxg 8SQJyVZmEAFanogILXMo04LwtKcZYUpPb1vKSRAEQRCKAWI+EZ+US3OL0vNaUGkwofQREVoB4IZH iFLjz8XWRxQ3vSAIglAocLsjMv2WTl5DgEqCbfkiIrSCwCWPZRThKaddEARBKDYQnIhPEpGkK1L5 IyK0QkGQ8sAKKpeAIAiCUChs6SVEpwjPykJEqCAIgiAIgpB3pE6oIAiCIAiCkHdEhAqCIAiCIAh5 R0SoIAiCIAiCkHdEhAqCIAiCIAh5R0SoIAiCIAiCkHdEhAqCIAiCIAh5R0SoIAiCIAiCkHdEhAqC IAiCIAh5R0SoIAiCIAiCkHdEhAqCIAiCIAh5p37YG1VVVerSSy9VS5Ysyef+CIIgCIIgCGXMypUr 1e233x4uQl999VU1evRodeKJJ+ZzvwRBEARBEIQy5s0331T33ntvuAiF9ddfX/Xt2zdf+yQIgiAI giCUORMnTjR/JSZUEARBEARByDsiQgVBEARBEIS8IyJUEARBEARByDsiQgVBEARBEIS8IyJUEARB EARByDslLUIXL16sli5dWujdEARBEARBEGKSkQi9//771SeffJKtfYnN1VdfrR566KGCbV8QBEEQ BEFIj4xE6Ouvv66++uqrbO2LIAiCIAiCUCEkLVafDsuWLVO33nqr+vLLL9Wuu+6qTj31VFWvXj01 ZcoUNWjQIPXrr7+qtdZaS+2yyy7qqKOOMq/TuonP0Z2pR48e6ueffzaV9DfffHP19NNPq5133tms Bz7++GN1zz33qA033FD99ttvie1OnjzZrGP77bfP9lcSBEEQBEEQskzWY0IRkvXr11cXXXSRevbZ Z9Xdd99tXj/44INVnz59jPB88sknjbBctGiR2n///dWRRx5pXufvd999Z0To5Zdfrl555RUjPm+6 6SY1cuRI9eOPP6oDDjhAHXjggapdu3bq0UcfTWyXdSJcBUEQBEEQhOInq5bQ33//Xb344ovqmGOO UdOnT1fbbbedeu2111S/fv3U3LlzVbdu3YxA9TxPbbDBBmrYsGFqo402MmIU2rZtq9577z21++67 m5ahWEihd+/easKECWrOnDlqzz33NP8D8aisC6688spsfhVBEARBELLIkiVL1IoVK9Q666yT823N nj1btWrVymgOoXjJqiWUCwywZn777bfGWvl///d/arXVVlMdOnRQPXv2VLvttps688wz1eqrr27E J8KVZXn87W9/M256cC+cRo0aqeXLl5ts+LXXXjvxOuvA1S8IgiAIQjT+/e9/q88//7zWaxiMCH97 4okn1J133pmT7T744IPq5ptvrvP6JZdcEujJHDt2rNELCMq44Fn94osv6ryey+8nxCerU4T11lvP zHD22GMPteWWW5rXVq5cqebNm2diQclkb9Gihdpkk03Me1hKufBwuVsLKWA1DWKLLbYwFw/rrKqq Uh988IHabLPNzHuzZs0ygnbTTTfN5lcSBEEQhLKC8ZhQubvuuivx2sCBA9XZZ59txlTyK4LA+3je eeept99+O6v7gzDFO7rffvupNm3aJF4/99xzjVHrsMMOq/V6Juywww6h30/IPxlZQps2barOOuss Iyp5kDSEqDzkkEPMxYRV87HHHlPrrruuWZ6EpQsvvFB17NjRuN2JC+3Vq5cRo8R67rjjjsbljnXT tXBiSeV/rKidOnUyy//1r38127TL8YP617/+lcnXEQRBEISyh5A5cjYw6AD5FuPHj1d/+ctfzBg8 Y8YM8/ott9yi3n33XXXssccao8+NN96ovv76a+PhJBn4jTfeUMOHD0+sl1wOanfjch88eLDZDpZM u74wGMePO+44U/bR8tlnn6kGDRqYBGXLO++8o/r27av+/ve/q//973+J19n3f/zjH+Y9QvcsbLd/ //7qggsuMJ5U8H8/rKVnnHFGrWXg+eefNzkuF198sTk+Qo7wQnjhhRe83r17h72dkgULFni///67 ef7SSy95Bx98cOK9a6+91rv00ksT/7Mcy7ssX7488Vz/ULxVq1Yl/v/555/NX16zr7OMvvDT3l9B EAShctCiA9ebedx8882F3p3YHHHEEYn9HzZsWOzPd+vWzXvttdfM83vvvdc77bTTzPNBgwZ5AwYM MM933nlnr0ePHt6IESM8Ldy8++67z9tuu+28UaNGeUuWLPFuuOEG74orrkiss02bNmYs//bbb707 77zTmzJlilnfPvvsU2fdLlpselOnTvU23nhjM5bDqaee6g0dOtTbe++9vbffftu8duutt3qffvqp N3bsWK9ly5ZmW/PmzfM22GAD78EHH/TeeustT4vfxL7vtNNOnhaunhaTnhaagd9v9913r7PMM888 42lB7mnB7V1//fXeHnvsEfv4CsnRExlzLnLWMQnXPLMY6NKli/rmm29MQhEZ8sxgbMklYDmWd3Fj Qq0l1LLmmmuav67FlGWIERUEQRCEVHg14V+VClZKqsrAU089Zf4PAg8jnkctENW2226bKLHYrFmz 0HVTQpGEZMooknzM+J+K9u3bm9wRLXhNvghWVryqLoQLENJH6J0WoSYMj1hWSjOecMIJJnEZi6jl P//5j0l0Rm+4FlIXQgH8y1B5h3KR1EEnDHD06NFmm0L2yUvaGLEemMvnz59vhKJ1zwuCIOQbXG5M fNdYY41C70okomYUM3AzKWcCz+Dvn9hnuj1i+xs3bmxcuKtWrUp7/UJxgEFom222Uddee60RidT1 DiKZ2OR641rwM3PmTHXQQQcZ8co1E7RMEKeccor673//a+qHU7axYcOGtd4/4ogjTKIyIXnEtTKR wFWOeA3CGqb4TJiIDFqG3xL3CRKmgdhZa1QTsktee8czcxEBKgjFATdxYqH8A8T1119vYqYYmIj5 JjHAQpzU+++/b55jRSHeyj78lgYGEtZv33/kkUdy/6Ui4Lb7Pe2009TEiROzst4xY8aYmPejjz7a lIxzj1smhGUUuzAYs10SLrDoLFiwwGQ/09jDJUpmcNj2Lr30UvXwww+b2H5i6YTS5k9/+pOxIJ5+ +unq8MMPj1RpBqG2cOHCxP9MRKyVk6QlGzuJNZMEoBtuuMHU9Y7KoYceamJPaWxDfKfLTz/9ZKyj XIPEpCJuYauttlKjRo0yEycIs3jGAYsvWgWrKg+srFKJJzfkVYQKglA8kFhI0gEPyw8//GAGALwX iJghQ4YYF5iFmz1eDXjuuedU8+bNTSMKHnzGhYB/kgvs+127ds3PF4sBlhVch5mC2+/kk082A+85 55yjWrdubRIz8sXQoUONqKBTHYM11UkQB6+++mqt5dg/EjzjwkSF84l7FLGN4BVKHxKOmFjy1+IP c+NhIUkIsUeIHWKR3zX3CZKKmbxSlpHPkuBEBj2WUCY0TZo0qbNuFxt+h8jl2kJY2io6LM8+YJnH WksoAE1u8GTw3p///Ge19dZbm88gHu017+67+zzs+7nPSUZiPVT64bsgQoXcIFVcBaGCsXFhxFIB AxJuNOsGo2oFogPrJ9Us/ODO4yYdBt4P//s0srjjjjuM1WTvvfdOxKJhXUOo4o6zdQMZwHCD0diC GsEsT/btddddZx5kyz7++OOmZBt1ARmcbNtf3Igffvihuu+++0Lb/TKAEvOFmGb7DGhsn0Hzsssu M6+HtRy2YPG85pprjEhDfAKDssX/vfgefH+szbgVKXnDd2S5vfbay4hXJggDBgwwxw9wE9rMZKy3 NP5w4ZhwfnAhEt8WBpnBuNs5b2Hf1+WZZ54x20YQYNXmOO+7777m80Lpg7hCgLquZrLZbbwsvy83 1wKROGnSJOOuJjYT+J/fHL83QjVsyB2/m19++cX8ZmwWvrtuFz5v4bfmLkMOid2Hl156yWyb65SJ kX0d6yi/T0SkFbzuvmPx5f9k389dhiL3TLjZLwSrzUMRso9YQksEBg4sHMUOP26CxSsVBBY3dPtA 4OT6PSsg0wGrA6VarCvLn6CAhQKxd/755wd+npq+CCMGHDvQuDAI8T4PBg8eWBcQTAwGt912m3rg gQfMsggeLBCUakGo4d6jlBtQwxBrI2BhIWEAsLTiTkQgnXTSScaSy8BxxRVXqHHjxhk3Y7J2v2yT WEf7HDcf1lE+gyiDoJbDLgj0nXbayQhQRJr9vtY1GfS9EIrE4vH9GEDtclhREQZMAmjeYcFFjqUZ 0cgkwR9CgRUWNzzvsa0wOF42nCLs+1rofod1i0kE3+/44483sXfUbhTKB3+so7U8QlCyL+9bAWqx caPu8qwDAeq+7q7bxf2cfxn/PvAbYhn/60zcrAAN+lzQPoQt434vEaC5JSNLKBYIZg7WlI4pnFk6 EJfhuqMK+d5HH31k3Aj2B5EJdJpgIMCCYSE7D5FIVyeEIvFv2YZB6Pvvv69TCxXRwIBHvJaFeDQG Ihszk08QJdR+TZYNyUA+cuTIxP/E3OBOyQUIKuKSaAmbD7AYYZWyuK6nXL1HnFS6IMooTv3mm28a Vy21//yWNMQXcYB0U/HDPcCeS4L3rSXQwu+SOElAVGGFxKVG8WngOsUSh4AErm+7fQQVNQixUCDK CAMg5pHXeA8IFSAJAmukzZbFLc1vndhHBhVc1WHtfv1gecXKyPawWiL2gloOuxB/ae8tU6dONbUS 6fBCcW2Env978TpinN8tos79rWDtZFudO3c26+SeAohMRCxQzxDRSGvjTPF/XwsWZCxQb731liQg CYKQMzISoczamRXbWZA70POaKywK+d7LL79sBgPESKYUW6cJvhdB4GwbdyQgWEkoSSZC6SxFTE82 BrK4EEtIOQ5rwfMP6pmAqOHc2zIdDNz5/o7Jsihz8V6mpcmsS952JvFbKhC9TL4QijShcOGaThYf iAUNS6sFa6abnIhbzXWPu5m4WC55sA4mfgg8Jr1M+nB/Q1C2LBAvZo9LnHa//kxZt+Uw58C2HHZh gnv77bcbwUozDX6TfGc3Ecv9XrzO/QPrLNm3NO/wb5/fLs+thdotWcd7VpxmSlj2ML9JrNtcEyJC BUHIFRmJUOp/EcOEBcUPNcVwkwWR7/e4aWerxhcDNjFhxHRxA3c7TeDmsvFWDJa4IgluZpAhvsR2 mmA/ubmzT3ZQx3qCgMIigWWI5RmsGeSxViWDdfJZ3Ib+khYMJAyQiGO6P1D7DKsW5bL4DOeP99hn ziMDPfF31qXHZ4mzQzRiGUFAcjzJXMTa7I+/++c//5nYNuviu2Dd8eOPJcSaF3Q8cPGyfgZ6LHG4 Qm2NWbaNOMKSxD5ihUOg4D4hQQMQG7b0SLJYxFSxceUMk5errrrKHJ+wLGyOIZM8LO+4utMFsUiM Jueaa/WVV14JjDUFJlacP2I52T+uAc439x2SFGy2LOcTUcbkK4hk7X5TEdZy2IXJFNcNHggy7xGu VjwGQbIX34drkLhP1yrL9yHpAzc9kynXvZhPOMacJ0Q+niRXxAuCIGSLjGJCccWRkVZJYEFE8OKm AkQZ4g8riRtvRcIDsV0IO5IUGHCwyOHaxB1HvBqizcIAissVlzsg5tiWDSlIBpYiXJb+0iu4/BBn iAy2y1/EL/tC/AwuTT7LIGvbpWHBYTC14QXEAxKkzesMnOwPwpV1sK/++DsL77H+sKQVfyxh2PFg /QhSxAri86abbkq4fwmoZztY5Dn2ZAQjOEgC4fvy4PwgVlLFIiaLjSt3uC65Rjkn7oTSnxnLsWdi ZS3s/vf9BL2PS5rzwCSEc+G664OWJxsbEdS9e3e1zz77mEx+W8A6LFvWn4GbrN1vWHasfR7WctiF ySjeFr4LE0Zc6VyT9tr3fy8mP1jqSfDhM+57TF75PKELTADB/33CMoz98BtnEmzbKiOQo2QG22U4 3uyrv1SOIAhC1ghrqZRp285iQgsY7/nnn8/a+mgddvLJJ5vntBTTosg8d9uB8bx///6Jz9BmrGfP non/w9qdAe1Hv/jiC9M6rEOHDnXW7aKFlqcHQNNeTItFTwsub6ONNvK0mPL0JMHr1auXaZvKgxZl vAZ6sPS0gDPPZ86c6bVt2zbxfTjvTz/9tGmVZq8B/rrHUA9Opk3a7NmzvdatWydaptr/9UDq6QEw 8Pixz1ooelqkmsecOXNCjwfro5WbRQ/uph3cb7/95tWvX99bvHhxrXXTik8P3on/tWgw7eYefvhh s88WLU49LQLMc44L5wf0JKIiW7SFtb31v+a2z43SJjdsmaqqKk8L2kjLui183eeWhQsXmv1yPx+0 XFC7X/cz/u3zf6qWw344jly3qY6TnmB5y5YtS3wGuA71ZMzTE6bEa3Z/3f+DvltU3HUFfd9sb6+Y Of/88yu6baeeVHmnn366eTBW0f4yLv369fP0RCr25wTBtu2siBJNzOizSTF2msC9iOWPuFdr6cD6 h9vTdn3AuoLVyA9uP6xKWJywBGLZ1JMQYwWyCSAUKHY7qOCutskybvwdsN+UwsGyicU0CH8sYdjx ADcejtg1LLTsD/sXNdnMLu/uvxuLGKWzRjkTZtH0xz+6Frgosahhy+CK94eOhC3rnn/3ucW6it3P By3ntvsN2mZQpiyuccJCSGrCskjiEwXfw+A4+mMog76X62Z3LZDgDwXxWz6DvltU3HUlyx7O1vZK BSzspVaM3FaISBdyGbCU2/OeTghSturs5ppC5kAIycno7oKoIaMSV1clka1OE59++ql5HtZpgtdt +ZYoIB4Rx1Zc4X4knhI3ts3stSAA2B9c7YDY5FwiqHG3k2jFoG1L89giwCRoEEOK29DGZvphUCWG jqxb3P3J6kimOh5hsN+48Yn9xA2PW5OYRfu9/MSJRRQESz5bDhM6kqo1p5AbCKfgUWkQz+xORriv M8nnHuzWpCUchXh94rKB8YmSXf46u249XAwjhJQQKsVzQlQwzIRtw+YXIBiJ/8cQw3hDmTA+R1iW NexQzxhDCWMxYxS/y2T77uZASKOF4iKjmFDKubilY4oVxM3YsWOzus5UnSb81oRcdZpwY7n4IVKq ibg9YlQRWZRLQoCRidgAS4oAACAASURBVMu2KFYN1tqKNRfYH5IibLwosW/cWGyxbNbL+yQ28R24 ARG7FhavRhIHpXEoLUPtxrB9toQdD//6+Zyt80aSEfvLPlENAJGN1ZtixyQauduKGouYKs5RqEzy 0XIYK02mlQ6E6KRK+Cwl0v0uVGrgYXMcuAdTUcRfk5YcAQQiIOgQl/wm/HV23Xq4f//73814w32W MclWbwnbho3/Z/yjEgdx+hi6sLYSr05+ApCfQDlGhC6eMJuLELZefw6EUFzU87zggnkUKsZyxN8w sNhRrigoO76YoEYfPyJ/jc1MYdblls7hUPJAxLjP3ffdThPg7zQB/HDcThO8HrQ+cD/nbscVbuwn ZWr8bkJew9ppP+9+H/aB9fjXzXHkB+2+jrh2XXbu//73wvY52fFw18F+ucKU5dgn95hiqWU5wgT8 28ISyvfkO4TtT7L9EwShPOA+QOUFOleVKtwHSdpzk0Kjwj0OiyZ/SbalPBsJmySv2ioZuNox4jBW YNAgXIwxHwMUyxLehSDFMMFzDAhM+PHGIVJJCmVM4VgjTBGsJMEGbYNlMBDY80F5NrxthMJQ9YTy blir6S+PMQErJ58hDI3xkrJnQetlckdiHp3frOdPKDwYjqi6U/7BPjkkqNNEmCXUvpZJp4kgS2hY Rwv/fgbV+vPHALnfJ8waGBSD6ReZqWL4kgm8oOPhrsO/XyznP6ZufVT/tqLEIooAFYTyh/sdnbYq Gep8++93QTVpcZFjYaT6CyX8qJQRhL1/Y2xhom/HFP5iFLB1tMPq3rqvc5+2+8Zz25GNdSOKba4D 4tNuJ1f1dIXckXHHJOrJCYIgCIJQvuAaJ74SayZWyGQQqwm4yAkHI+kVY0E2Gh+Qn0BojG1IgjhN lZfhz4EQioeMRKjb6aOYwVUQ1slIEARBECoNLJMkc1oBR6xlspq0uL2pakLcpiVZrVnC+XCd4xZH AJKs6l+n/3//6/7atUDcKaEDtBrGAorgJa412XptDsQpp5xScXWgi52MYkIFQRAEQSgP/LkHyeL9 wY2fD4ulx33udtxKtg33eapYffIHEJm29FqqfffnQAiFxcaEZpQGjHl+0qRJ2donoUjhx7xgwYJC 70ZkSFQiIL5YwApPyREL5ae4KdvOOgT7C4IgFBrX+gjJ4v0hWZ1di7/la7JtuM9TxeoTf2oFaJR9 JwdCBGjxISWaYkLvauJcXMjcQ2RQuB51T4aehTJOtpUnLSoJhO/fv7954EJwSfV+oSCex5bXoFC3 vz1oupCtSPkO6rbRKtQ9bplAy8+wHuj5hmNHTVlb3gSIZyKoHk8DdSGpySoIgiAIlUZFFESkMDmF oLMBfdYpUeFC5wniaxCnQ4YMqSUqRo0aZYpdwxdffGEK91ITkweFfV1SvV8MUI6LXtyZQrkNaoiy PqyBrVu3NrXfygmEJgWeKeJMoX8XZvGcX+KryDYVBEEQhEpDSjTF5JhjjjH10CiUjmkf1yoikyLz xM9SFB4LH9bPoI481E5L1kEo6H1qW4Z1nqA+GgIHaxv7gQWQYG2EDW4QCvVirabQLw8+Q1WDqqoq Y93cf//9E50qWCetNrHQET9DDTnc2m6BX4oP4+6mThviiqLAdMigqD4B37YdJu8RqoGlj1IdrMuC xfOaa64xohvxCRSpt/i/FwHlHG/qx5FpSXA835Ft0AED8co2BgwYkCiuzzG78cYba3XOyDd8P4pA c6ySYcuLEGhPoearr75a3EaCIAhC2ZORJRQxs9VWW2VrX0oC6qVR2Nd2mMDCitXSrVOG2LPtLv0g rhBGPAjYjvJ+ss4TJ5xwgrEmItZoV/bee++Z97DOYm0EhI3tM4yQpd0obn+6UtDNCBFKy89x48Yl ih7TeYIgb1zlrvudz9vwAjpk0NWCjhaIcZt1SMFiLMZY+ejJ7L9G+DzCFgGKGLbf17bqDPpeFECm uxPfz7YyZfsIcpalBAcxyhZc8v7OGfmGbacSoC4IatqQ0glEEARBEModKdGUBlhDn3zySWOxfOqp p+qUfKCUBfXU3GQUi9suEtHk73Thfx9LJV0ibOcJBC4WTsQo9OnTx4hUQGwNHz7cCDdEGZZRvjev 8R4QKkAiDNZIrIazZs0yJS4oQo/YxAJHgV9ENiKPbdIG0y3L4YLlFSsj27PiidhYhCk9hXE5u92J gCQnW/R+6tSpZt2zZ89Wbdq0Mdv0fy/WgSBHqHbo0MGs34K1k+2T5MM6bXFiXP0IabjggguMwKVz Rj6YOHGisQan00kMKzTWbmKLEdGZwnGlNl5Q04BcUqjtxgWrPol3xdizvVD7xsSQ3xgT7kaNGuV1 24IgVBYVERNKP3ErwrIBAumFF15Q33//vblZ77rrrrXeJ0uPBCZEGYOICxZAkpV4BLVa87+fqvOE 7VABtDn73//+ZwQkVjhiN3FtkziFmAXEIX3gsUZirbQVulindQFjGWXwsdtM5hq277E8LnDA8onL nNCE6dOnG3e6y+abb26srlgnO3XqZIQnwtu1VrrfiwQtYivpXU+cpbuc3T7dMXhuj3chO2dwHtId vCnojHWXcxYHwi9IZsMyTStChARgNScsIF2YrHC9xyVsu9T4Y1Jk4XwSalEooiSx4TUIqWQXiXSP Ydi+Zbo/yaAMDvcgwmVGjBiRk20IgiBYpERTGmA5xNqIWxuhGNStAWFKIWAspZluC2xGfrLOE1jQ 2BdECFZahBtimK5WWFOwpiLkcGcjVtz2li4siwDFogfEKsYBAUVfYay2CA4Ersuee+5pYkjJiLeC 0i/WXUj2uuqqq4yoR8C68H0AN33btm3NegvJtGnTzEAexw3vh0kCIQZxoNA0WfcIf2Jpjz/++LS3 78I6rUU6G2BVp2C0LaFFmAkTrmQQjsE1UyhoC5hJOEe2j2GU/UHss1xcmDRy/Q4dOtSEGQmCIOQS KdGUJscee6wZPPlrcTtGAP11cZtbsed/30/Q+27nCeIicTOHdZ6AQw45xAjV7t27GwswMZm8BohL xDGJVVhGEYd2Hf71IF6xYGJNJW40qHNFWLeMHj16mKQhkpFwlfvDFbBYEmJA6EG7du0S9TJtQpZ/ fwh/oEXbvvvuaz7jHiMsvXyeCRHJW0GfD/p+uYLY20LFSW+55ZYmEY1jj0XcQkIXVlLCEkg4wwJO zLK1phGq8M9//tM8p5oDApjlCZtgcuD+dogDR+ASQ8x6cLsjeIgd5nynsjgzucEqSHxvEPymTjzx RLNO9gtPgE0wY+LEPYeQErsdLMDWasy9CAEPWOW5hgnJ4Lth3QeuF669M888U91///21to1FnhJk UUCsYe3nWHFMLJRZ49jxOvtL2E62j6GF72TPNcfMlgGz3g+EpA3twXPyr3/9y2yXREJ77vk89wju Y+wj+zR//nyz70wUWFffvn2NiHavKUJ9+E3yHr9J+935HK998MEHkb6DIAgVjhfCCy+84PXu3Tvs bYMe/L311lvP04Nc4jUtGDwtQhKPYnivQ4cO3j333JP0u6SDHujqvKYHnFr/r1q1Kun7qT7vsnDh wjrrXrlyZZ3l9CAW+NxdD591txW0HN9PD1619svdpn9f7f/Nmzf39OBqnmtR5nXs2DH0O7EuLXZq Haeg76UFRmJf7Hs777yzN27cOE8LlVrL+z8f9N1yxeeff+5pkZHROth/PYjH+sxRRx2V2O5LL73k bbfdduY5x2innXbytJjwtFDxzjjjDPM672vBaZ7fddddnhYU5jhtuOGG5px9+OGH3qxZszwtkMzv Bx555BGvS5cu3ttvv+3piZHZxxdffNEbMWKE980333j77befed1ul3Pjp2nTpp6e1HgtW7b0pk6d 6o0ZM8bba6+9zHtaBHp6IuJpweldf/313h577GH2iXXa/V2yZInXs2dP78033zSf0ZMcT0+ozHMt Lr3DDjvMPNdC1zv33HO9L774wtPi29MTKvP6oEGDvHXXXdcbMmSI9/HHH5v/BwwYYO53fDeuRT96 0lPrWuf307ZtW+/pp58269hxxx09LWjNe506dfJGjhzpDR8+3HzHTI6h3bew/WG9eiJpzuno0aO9 TTfd1NOC0vv+++/NvU9P7Dw9IU0cDz0p9rRg9PTk1Bs8eHDiPHGPZPszZ840x0VPZsz6+P3feuut np7Ee1qgmu/D8Zk3b563wQYbeA8++KD31ltvmXVxTDbbbDPvvffe89544w1zHX377bfJL1pBECoW PWE197iMsgZw05LNa8viANYW1zpaLO9hncs2NmbSxR8/6bfApSq9k+z9oM4TQRa+sA4U/vW42wpa zv1+dll3m2EdLazVFssc14e1sgWBVdMfWhD0vVw3u2uVBcpCJft8PpNjtt12W/PIhLlz55owj7jF +7G8E4qB5fqBBx5IvE6VBH/yGJatxx57zFjFsZhR0YDzRywr7nwqJxCLjLvcQmgJFkQtDs3DEpY0 FgbuaX6bWP+0UEy8/uijj5rfKdZkktpsuAbHk6Qz9hVSJeCRFBY1mQ/LL2XJbCx1UJiLHy3yjUeB /tXA98C6iPWTc8ex5prT91mTXKZFW9aPoYVkO+sB6N27t7FK4r0gyY/wFJLjsApTPg6PAhZcypxh KcWSCVhI+b0Cx5R4bFteLiiRkRAslrMhI4RKYEEm7MdW9GDbhMhg3RUEQQgjo9GZ2ER/LUxuTmEU 03tC7kAAMeDj+iWmNVc1LymPVYxZzZlCLG46lScI1UBo+I9JUPIY7lfEyEUXXWTcwfZ3jCjDjU08 M+5pG5MMuMcRSS4kjRF7SCwqSUZR9xshxiTl448/TriGETBcM6wHWG/QRI/ELx4k0JCAR4UF64Im oSZOMh8gFHElI7SiiNCFCxcaIW2xtXGZHHF8evbsabbJd/Rf+9k8huBOsDi/Qa58suzBTmoIgXGT Bf3Hw4VERtbLtWITGYO+A8eca8ueO8Jj7KRBEAQhjOKunyKULPT0dfv65oJ8lVyKCyKC6gmIrHQg KzmdRByEUVRRjlULi92pp55qhCsgYPg8GdlYHrEsYhG1EOuK2ENkIRZ52KQxYlFpFuBFzNpG2Nh6 ulg9AYsn38Fa6EiQwaLNsgg/i5uAx7ZJVnMT8Ow+kMyHeEqWzAd8jrhrBBfL+j0OflgnyyO6sMTS kQ0Rj8UToUbMNtvbZJNN6nw2m8cwGeyXPWZ8b44Llld7TXJsU2ETGRGdXBtk69vvwHck1hURjPWV Y0JMKteTtQILgiCkoiJKNAlCPsHdaxsFxAVhw4PkljiEJb2FJY8BySyvv/56osg/JccQpri0qXFL mSU3tAG3M4k2WCCxciFSwpLGwvbHtdxxnBDsNtSC9RMKgFgiSc26exGZCB7q/dqEl2QJeHGS+ez/ rIPvQlKNH6yoWP4QlWyPdbKPVEBAfNravtY6So1g/ue74ZJO9xiGhdsEhcbYY27/5/tQiQBLMSAQ OT5UzGC7hGIEnSd3HWGJjKwTSzlilIkD54zj0KtXLyNGsejSAc0N5RAEQQiinhcyZSWGiBs5f4Xa cMhwOxH3RLwUHZSyAVYR6zoMA2sL2fZYMoiXi+I+FPILFigECjGObtxfLuF6CAp78L/u/o9AIjaT NrPu+wgjrisrRqzFy8L7WAuteLGuaCyWXJM8j7o/wO/JFVNkffO/a0lnGVy+WBgt7n7599HCZ1zL JuvhYffd/3/YeoJgsoCF0DZjIA6VjHvCRABLL8eGSgDpHEP/vgUdQ3e9fM4VpoQf8Jp7PyFemPAB G6oQdD78+8ox5DOsy10Wqy/75sZrczy4j8l9KXeQ5oCDgiIc553HJKnQeyQI8aEhS/v27cUSmg50 FiL+jJg520IyG0W46RluLTZhUIKGbWJtobyKUHwglGiVSskbG4MZBSxK6RIWdxuWPAZYx2xXKfd9 whySJXZhIXOFESLEFue3r0fdH/Bb+4hR9IdysIwrQP37FSYcg5L53H33/x8niQ2Xt9sNDEstSUUk CGHlJdkJ93TQeqMcQ/++WcISCl0rJjBZ9U9oEYdBCYcu/n3lGLJe/7J8d39dXtYtAjS3zJ1LgptS V1xB6A8xxUymcrtNLPtMsvIF3gK8QfZhy4DlgrvvvtsYDDKF8B9ChIoNkoPtfagYEREaE6wCZCGT 9UnmKMkHkE4R7rhgiaALDW41rFeSeVq8ELfIjduNzQsDyxMJQsQG5hPiKG38pZA5JDiRqY81lIkq Xcmy5SURBAs5YSNHKkVX4NmzlaIvRffuSo0enZvtUWcbUUi1ilSk2yTBD94ErO80TOCRjRbGFn/z C0JOCB/JFCagUQwJtMOlBnMucbdBpQxaiRcrkpgUE0Qo8U+4sQjyt5YWtwh3kIUSqwg/LKwTxJ3Z Mj5kBzN72nDDDY1LzoIrDWsnsxh+JMxksKzRso+WjMSPZdKVR8gflP6hDBJJPpxLNxuZMmcIQZJz ohZKzxbFmthV6rhl4gQhF1B4gP4HWEGpuvbRR4gp2uUqdeONVEDI3rao3kDiHGMb45FNbmPCNXjw YDMW0qSAUA88hHgxKAWGsQZPECXEPvvsMzMxpykJ9z+S86iggmijig1x234Y32wDE2Ayz+TONj+h S1yrVq1MzDJjLrHKlErDOs8ytnQfzSFIHqQqRL9+/Wo1v7CQlEgMd7L9JdQEActnTzvtNBM/H0bY 8ozpNINArO+1117GkERlDrbJmE+cPmXqCEnCU8W2KSGHtZbvRvUSYs9plNGu5iT7zwNx/nYb1hPC d7Hx4UFaxG6P+HvyAdA4+bKeigiNCSceQQFuf2+slLjK6YWOUHThx0ItPmaIuOnJguUiR7gSxM/J x4I6cOBA88MFLkbce1jIuOC4wBArJDSAW19RKG6wXHNTIOucGSldcWj5CsQUYzkjW1oQhMLyxRfE 6xIrW/2gm7B9Hve1fH2eOQ9xojhchg1T6oUXCLVRyhdpkxaMOwgWrKCUQsNtTawz9zPKfPEcaz/x 5SQCMplGGJLUCNTjxYqJVwgxy+tU/8B1TTk1Ww4uCGruIuAIBSF2EAMQgtKKUAQZ4goRSmgc+Ss2 pI1lEHbsL8KP7bD/iGPGX2oRW+sg4y9Cj3twsv1lDOceTi4IyZuUPAvrgBi2PNtmjOd/vjffiaQ/ yhpiXGJ7JF+yXrqXkTRKzWiSRoGa24hIRC3COug8uNsg7AcLLeIdr22YFkG/XH755cZQgvjkOCBI qTmca0SEZpGwIty4WflRUDiaBzMNLiBmIrgFiCEDZk0kIqQqLi2UFtxQOHc8OL9uLUe3BJIgCIUF 8TZ+PLGy1Q/CY+3zuK+5rzdsmNnnw17jVoI11IZM4mS5/npaHWfneOCpwVKJVZHxiOoKWEQZj/wN C8BtkhCnaQQl7Xgdy52tcoHgHEncgVKRXPwIPn9TjqDmEMReu80vEJmQan+Jn7cx9IzzlC5L5k0K Wp6KEoz7uP8R62FNHqjAwT4iHm0sNlZcGkWwvG1oEXYe3G1gTbWx4mFaBDEa1vgi14gIzTJBRbhx I7j1G/lB08mJmY+bOMHFxsWSqri0ULpwfrlJCoJQfNSMwSUBFs+zz1ZqxgzuK0qhQ3DFZzMaBGsg pdCs6xnrGSIpqGGBnzhNI4hfZNxzX8MD6OY9YJDBYhdGUFOOKPsZdX/dhD1EXlBjCJcoyydr8sC2 7XfCY4ZlEx3AuuxxiPP9IEyL+Pc3rPFFLpDEpCxji3DTBtFia+kBFzTxIsR94LoYPXq0cXlwIdga iG5xaWtBy8eMRBAEQShucMEffHD1AwFKY8BRo5Sil0A2BSgluLAS0sENdzUP3N0IUyx6uMNJqgSb ve4mYtpua7imIVnTCJJ68R4mi0PE04hbmrGSfRsdIRPLNocASofRdtbf/MISZ38zwT1GthMZ35tx Hjc7lk4/nAcsl4Ry0S3O/X6pzoNLmBYpJGIJzRL+Itw2QBqIFSVeBtcGP4Tjjz9ede7c2VhKiSHl QkR0Ugzbms1tcWliYfgMpng+JwiCIFQuOMpeeYXSWUpdc41SRGnlojMyrmncwjbBB3DJYxUlUcY2 LEDU4Vbnf5okkM1O8wi6bdmmESTeIorCmkb4CWp2gXsaSyHjJYIUD6FdR1hTDkLjyLsgjhK3OjGi eCpt8wvc3ba0mdvkItX+Bu2/u91ky1uLJsIb17ht8sD4Tx1wqur4y62RoEW4AZ9DGFtt4TaOcM+D uw2e23WFaRHib8MaX+QaKVafJaIU4cbdgLnfvxw/DjIK7alwP+MvLi0IgiBUNtOnV4tQX+ncrBLW LMFtZhDUsCCoSUKqphF+wppdAOMl46gdX3kka8oB/uYQbvMLf5OHKPsb1tTCbjfV8ghAxny7j0FN HvyfYT/57ohx//cLOg/uNvzLB2mRZI0vcoEtVi+W0CwRpQg3F08QtjB30AmXws+CIAiCyyab5H4b YSLEFUZuswYLMYt+gppGJBM4YQIUgsbLZE05wI2DtJ+1zS+ChHCq/Q1rahHWUte/vGtdhqAmD/7P sJ9WQwQ1jvDjbsO/fJAW8Te+yBciQgUhDzATJcYIFxdxTcxcySQlgxGXiv8mKQiCIAjljiQmCUKO GTZsmElCoyAzM97u3bubuKnWrVurBx980HS3oCYcBYMFQRAEoVIQS6gg5BDiXgjQJ9HM1qlzoQwJ cb8s17VrV5OZ6e/HLQiCIAjliFhCBSFHUPiXlp10pAgSoBZigR577DGTeUrLR0SpIAiCIJQ7GVlC aedFBf6gQORihTZftL7KZ+CtUHnQho0+vsR+Rq1sQH9myoLQTo7+v9m6RsmEJPMx33GnhdpuXMhi pVg03UoEQRCE/JGRCKWfLAOt2+Wg2KHOGC7QFrmsbSFUNHTC4rdBcWe3TEkU6HNMzTdEKDX34kAr t0GDBpnnWFSpGUc/YeJOSYaiZ3A60Lnr/PPPV0OHDo31ubDt0kqPntDUrLNwvKjpV4gJLTXyqNVn W+EFQVwvtQ/JMuX80CaPunyCIAhC+mQkQqnKT99SKvmXCnQkEIRcQj9gfhNkvscF6yfdtm688cbY IpR40i+++MJ0NSETn+1/+eWXsffBD4WhbY/nbPDMM8+YDiB4JY488kjzGoKVuNhkIpQELrqKJevX nCueffZZ1bFjR7XrrruaQt2EWTz55JOx1lHI/RcEQShGxCctCFmGtmj7779/2p+nOwYiksLEcSG+ dM8991QXXnihsYKOGTPGvE5/YoQtGfqsG7CaTp48OfFZ3scqSBs4LH39+/c3PZ1xqY8dOzax3Pjx 402Lub59+5o2cbw/ePBgE9N65plnqhn0EkwBHhQsn7bHs8uUKVPM+6zftrLFQjx//nzzGUQ+ovXj jz82702dOrWWtZVY3GnTppnndEqha8gZZ5xhBDpQheCWW24xyWK858K6LrroosB9ptPK3nvv/f/t 3f/r3fP/x/Hz8QfIb75ttRkpbYSIaVh+EJnyLSSKkvKDobBEypeaRUkrTSRiUgujNAq1sR/U2Hpj ef/C9J4ahv0yhTmfc3nwePV4Pfd8ntc5r3Ne53Vee92vtfY6X57P5+P5POc8n7fn/X5/3O5pWevN x/Xpp59u3XzzzanMJzscaIfHjsvxePHFFw8bfxAEQRAiNAiGDpF2wQUXTHt5Zswsnb755ptpr0M9 pjpH0TcQbSeccELyJNXGTUcM9aqieiCQnn322VS/eeONN7YefPDBFP3U/YSw0i86v087ODZT3rdt 27aUcgcRZnu9ZBvOO++8lEXRNq9EpxICXoRUbaz/lQMQfwyZjd2yxk7cQZRSOt8+6zzyxBNPtI49 9tj0vDpb47nkkkvSOozV/rDE2rFjR+vaa6+d2LbXrN9NQDcI73xcHaMffvghiX43DTmyKypNyIuI 28/q+IMgCIKwaAqCobN3794k+AbB8sQN26Z+2L59exKJ6hvvu+++lEKW+hYBFa2D5/fv358eE0Qi eZs2bUoCSomNWkfR3Ntuuy0JJ+PIfPDBB0lUEVgQdcUdd9zR+vrrr1sLFy7sWltZsnbt2taKFSsm 1gVjtQ5iFCYLEboEr1S9EgMC88orr0y9lMECyz6ro1U64D3GbdKk3syEoH/2ccuWLUkQ6hhCAKvx tH8HDx5sXX755SnKSajXsW7dutaGDRuSe4GJZ7///nuKdmoJaPIZ8avTSo5g33DDDZPKGMrxB0EQ BCFCg2DoEBsiddXWbP1g+elM0iEQCaTjjjtu0uz6siWb9YoYEninnXZa68MPP0yTjqSMtZqTjidM ly5dmuo3rStDvC5ZsmTSNvfs2ZMifISgdYuy9oKaUNFOYjS3uCM+pbhFcWEiYV1U2dhFbZUWGJPI 5ubNm5MINRZwGihn5vs8iE2YMFa2sjPuBQsWpMlkRGodhLySg7xOpQvEbnY/8L/1ih5jLk3YDIIg mA0iHT8DSAuKjsyX7fYLS5zvv/9+ZMuNGjPTpa0HwfLW0y9EkChqr/ZOInUifIScqCtxSmSJMoqE vvfee5Peb2Y4kaoOFFLTopcijcSkCGU/EI9S/b67OOOMM5IAv/3221PdqVS6yChEaQnLDLGpbMBk Ien2jz/+OEU6V61alV4//fTTU0QXhKE6zfPPP792HI7Xyy+/nP6JqNZB4JaiNotzqXcQxFw3qj2g M9XxB0EQzHdChA6AtKMJD/6phxMRgvq7nCqcLi6+X331VV/LNG2XJY5UYYkauhyxGTXZEqcbZh6b GJP/ERO9LDcOGKM07XQhtO3rsmXL+lqOkKoTn6KMOdJYfXzNNdck8XTTTTelx+oizznnnCTwcoq+ fL9UtQgpMZpFnhpKAlAk1Pcvd3yqbrdunIQdqyYpbJHE5cuXp+NHjF5xxRUpfZ3LAXK09fHHH0+P tT6Vis/1lkoPRDuzeLderxOeZ511VuuWW25pnXnmmY3Hg4AUERbxrDYMqDu2lnnhhRdaV199deui iy5KyxGxTftepxDbCQAAEUtJREFUHX8QBMG8p93A5s2b26tWrWp6OXH22We3Oxewru8ZN4455pj2 L7/8MpR1dS7S7RUrVrQ/+uij9pNPPtnuXPza+/fvbz/zzDPtNWvWDLTuzsWz/dtvv/W1TNN2r7/+ +vbRRx/d7gi7iecWLlyYxtqNJUuWtH/88ce+xtALe/fuTevuhjE/9NBD6dj61xFmPS03U+zYsaN9 8cUX9/Ren11HoLX//vvvaW1r3bp17Y5gmtayf/3112HPGcehQ4cmHv/555+T3r9gwYL2d999N+k5 n3s5/nIZdERj++DBgxOPrf/AgQOTxlDd7lRjLPnjjz/aP/3002Hv85sol/e+cgx16zau6vPV/Skf V19rGnPJr7/+Oulx075Xxx8EwdT4DU91vRpHnEtm4ho6XZyX9uzZM9vDSNx1111Js0QkdECGYYkj 9cpAX/0bQ2xI8eUJDvPVEke9omPrX07JTjXuvK133nknRb42btyY6gsdP59RttARBbYP3vP888+n VD8sa9zGbwJP5+SX6hR9VgzWba/8DOsQyRORe/XVV7u+rw7pWtHE0si9H8o6x4yIXFN9qFS6Gk/f 3XIdvCzLSF65DEQeyx731m+yTzmG6nanGmOJqGhdWttxLZcvu1HZVt26jav6fHV/ysfV15rGXGJC UknTvlfHH4wR27ZJDZhtZxZaq/VvyclYs3Ztq3NiGvpqnffLUhzn0PXr1w99O73SS3bRdSRnzu65 554Rjaw7rv3ZMcN1tOoGMh18NvbR9Uk9fz+MYzYxROiQmK4ljtQhcSSVKE2nbgwmhOzbt29eW+LY V/WCxJzt9DJu23r44YdTelhNI9FI3F933XXpB5jLEm699dYkNpy4jDt/Lo67ZbxfnaTZ1SyT7Ath Zlsnnnhi12NNgDjOToT5BqQXdFpSc0gw+z6MAj3tBykdCIIjgq1b3T06uTF5pWharcoN81iyaZO7 9VaL88UQxaiAiYBBnmOgJObtt9/uusznn38+4ZYxGzz11FOpHlyZTr+16aNA3fyFF1448HrU8Cuj ogl8RsqmukEj8HseV2J2/IAMaokjKkKIuqMhdKoTJ+azJQ6Rm3881RnXTeNWm0csiuRCjaFJLu7+ RJAJf9FgkVKR5G+//TbVHzrOxB/sp7pINwQmvhCr6h/tQ6/+n+o5CVrHxrhMnukG8eyYOSY6Ho0K AjsI5j2d89BhkU/1vZwSxrnF84EDrdahQ62WrMvGja2Wm/zHHmu1BryJlWFgMSYw4hxdhSDlRmFy npa+zutltujOO+9MN/bOpdal7a1sgWuBc79ggH/OxTJpO3fuTG4Zbty5SphEKBOo5tz1r8R51XKu UVUENMpMQ6/rl7VwznYOlp0zXhFHWF7Ag3WdaLDMo2uG7KV9c3xM6pRRc73gsuFaS8CXnsCue65p tpOve5aRUXLdlCVh++Y113YBCddV66riOmc/Xb9kHwV37JPMn2AGbWHMZUZTYMhysIy5Iq5nro+z SYjQARnUEgfSwqKMvjQidERrZj5b4jgedSeaqcZdHntCMm/X34c6J+w8E1vkFIsWLUrHMpPf76RT V77QK37wTgxOEE4KUvzVCKexOI6i4T6b/J0IgmCEdM5ptciQjPONWunC4TogWql0SblUUarSL64p SoJcr5y3SnLmSFBBoMH5TUDCeVyDjZwtEhl1Ay4AIEop2OI9slS5EQd3DhlD5z9ZL+dJAQblaI91 xHS2iuPIAcEDbY2z60UVItj52/XG9aXX9Svncg0mQnOmT8qbQCSWZdW06nVOF30ktAU7BJuIYsfB NZHQFmxy3Fy7CPRs1UYIy/6J1lrGvgg4mMxIhBK79s01SjbPWEstUIfyvFNPPXVi/cQ/rWFSp4nN jrfIqWu0/QT9IDikjM3xsfwwIrTTJUTogGRLnF6pWuLA3Y8vhC9C9YtnFrIvqSgeceVL5weZLXH8 0F955ZWet++H4k7MNuGOTq2muyHrz7WRyJYy2Vy7aolz9913p0iaHxrybGnR1myJ03SXlS1xRBzd LU7VpaZK07il3KdCrSGxLBWtFSP88LtBkE7HXsdJywlPRMD2nIh9Xwh0AtprxLM7UzO4gyCYBdyE /it0Jli8+J/UfI92Z7NC5zrQ2rHDXb302D9lBMcfP5RVCzC4Fomi3XvvvRPPK4+SsSMkdXYj2IjW arbI9UJdqUCNa4m5Da4LnvNaP80eXPMELJwnicEmG7Q1a9akAMvizmdH0PW6foJyqkwf8Sj4QegK DJUZSIEk53oC0PiITNtU1kBQ1lHNuME6lYI5rvSB7dZhLCKrgkOrV69Oz7keE5iCK64zXFZcU8qM pmXcIBC8eT00RYjQOcp0LXHcWZaTcaTH1VyqATXRqFx3aYlDCPnRqHnxBfZDYSszHUscX9iqJY4f rjFItfii5mirOz93arbrzk1KI1viGHdpicOuRqTXhJ5siUMYdrPEsQwxXZ5YrN8J0N0qlB9kuyA0 jdt+VreV9708PgTsVVdd1TrppJPSWK3feMtjVf7tbtGNgB+1CECT32QdfvwiAT6zL774Ih0PJ0J3 6sYfKfEgmGXc+LPEe/31fx4vWvRPveU4C1A49xMhQxSfJa5HbtRNSs03+oSYc2bOQj333HOTJghm lHD5J8jgOibNLTCh9Ml53Xr6afbg5t1kVCKrSYSKYuZMVj/NJHrJ9Imk2ldRRvteZrXyupRVuU7n bXabhFiXcXOdl+aXXXWNKDN0Ja5Bx3c+77KhCfFqXdlruQwolZSZQu+vzrcYOU3T58OiqTeGYYmD n3/+ud35cTSud75Z4jRRfW/duKvHu6T62LLlPnV7v2M2zO9OEARjxv/+127/5z9++LM9kt7oXBdm go54m/h748aN7Y4AbHdu+tPj1atXtx999NGJ1/M5cteuXe2lS5dOWk9HwLY7ArT9ySefpPX4e+XK lek159PODXp7586d6fFnn33WPuWUU9LfVbvB/HjLli3tk08++TBLNBx11FGHna97XT8eeeSRND77 53q7ePHi9rJly9r79u1Lr9v/999/P/39wAMPtO+///709/Lly5OFX8Y2v/zyy/R3R2xPWPuV2yyX 8b/HYPlIexlzR0geto91+wn2VbSNazId4bh3bhzSax2x3N69e3f6u2pzaDzGNRuERdOQGIYlDtxx uStpWu98s8RpovreunGX76mOo/rYsuU+dXt/jt4GQXCEwvlC7dxcsdFqSNcOSnkOlbYW9cvXH+l5 ZVfKi5RR5YmuZbYoW+bJNokumsl96aWXJgtCz6GfZg/5sXWYUKp8rUr13N1vM4mpml/Yrsmrl112 WUph12XMoM5UBFOKW9OLukxcU8ZNhs3EJKV2UvUykFPtJ0RxlckphTCXQsQ3b6tsklHd72rmcDb4 P4q07gVFsz5A/zehLlGBsv/nCkQET8vZEhO5HiVSsME40blDTmUD/dwQzOXt9stcGWcQDAM18lWx QyqUgkXq2ePyWuY90uzl9TXPZ6j+XWKZ0mvXevzL4qz6uG49dWPudf0ZqekclFDn6j3lOtWxWkbA yOv+rtuu9Xjd+/Lr5Tary+THxqhuk4hUe6pmM0/i6nU/CWfbLt+jfMLn5Lny2HlfUxnfTKOWVTlc REJHjC9CCNBgtmgyOlZvm5sLTIfpGjHXbVdxvhNUdi+Ael92aLNFr8eH7Qmv3Jmm37a+u3fvnnUr lmDu0Ev2TB1k9VpWly3qJes1VbOHbtnFbmPudf2ZqTJ9osE5Y5mXr9uu9eT31WUqmzJuOWprfoJW yrzA+91P26m+p8xolsduHCKhcVsfBPMIzgxmfZoxasKZyVJlA4HpIhuSC/4HhWMAz1nRx2zEbHYs izET0pogsE0wyH6vs4GZwqIObFqq1mrDxCQEE9u6waRaWtTMXxMNWecEQTC+5AlJziFsH+dDd7WI hAbBPMOdNuHJU47lSYYJc46SOgny+HvttdcmXidec3clFh988LKRtRRSbh8rTScayLv19X9nG5sV m33pym02YQYqJ4G66Gddy9U865bbAk9AUcsNGzZMLMO6RX0WrFPEEsyezUZV18boOWO/1K+JRpQR R6ksM2f/W9OdRstenrMEaHnc6tryEtW26XjnhgxNY6k7nrmtb1ML39Kkmqeh2jLR0Ey39rpe41fM QSIIgtEiuuymcT4IUAwUCWUPoJCWRc5cQc1Ebo0ZBPOZ0ugYoqIK4rN5MlFD3BCb0jaEH887IoUg JfIYJKM0Yibc2Kmw8tJ5xPJ1RspNNiuQMuLrx4O2KkQJJJMf2JwRbWqkpLF466nblDoX/ROVlILm l6ebCwNpwtW4daZyLjC5gvAiHo2PwGVMXTWgBrFrfR7XtVVl4WJf+SWKVFoWRLHzpOPJNoZw9D7l EMZgm93GUnc8jc/7c0m/tJ3PRZo+dynLJtUmLShxUC5hDGxsfJbsZljNeK/PR32Yz9c27ec4GFkH QXBkM5AINUNOymwuoV6lyQA2CIYN0fXWW2+NdJvqf8wibaLO6BhampbmyYSW1qhbt25NAs+EPlFU s11F1bSWM1O0hHDiAMFrj0tCbshQZ6TcTYQid5AqeyN3a7kqemCf8kRJYs5YRfrsM5NsUUpm0iK5 fADNKOXdC1FDkVUiDNmAOkNUSoHXdbXyORsXEa9WTk0Yc22zgqtteQlH9WLOn7blfETA1o3FftYd z4zx1LXw5TdYmlQbU6796tZel3gfJyPrIAiObAYSodnGIAiCerQtHXUtnohft0krdUbHqDNPFg3T IYXYY9UiIkpYETKiiyJwoqIZXaXYjWX7sEyvRsolRBO7E0JN5BJTtVwtYZQtOmh2qXGKChKGBJrx 5bFmcu/mTNUsW03qp59+mnozl3ZqIOIIRQ0UoHkCYelY1bXllY7PLQNFNZvG0nQ8M9Np4dutve7Y GVkHQXBEExOTgmAGIdqyz9yoKMVMHaKFvdYbEdDqEUUTFc1DLSI/OpE/qWcRyYyom/S4SKlOKyJp Usn894g0ouall17qeV9EBwk1NZwikd1aruY2sxniTB0lYaZ1oCiidLPUPQhiNnQEt2UJyW4Tn6Ty iTVRXansEoIzRy6hs4uxi5oSrWVbXsvbB+UPxKUIrVa3dWOpO565BzRESZta+Fbb7mb6aa8bBEEw k4QIDYIZRDRNVGxcaBKfTebJ6h9XrlyZ0r1Z/Lzxxhtp0op9MwPcLOzsNWe59evXp9SuVDG/OwI0 GynLnpRGynWtb6vPqW0kIHPktqnlqhIEhtMm/9imiKeoa+5lrRaV2BLFhVQ7Mbts2bIk3JXq5AlB 1THkx0oWiEMp7OwqQJhKt4t2lm1eRWk9T1yWbXnVzhqL100kIrBllOrG0nQ883iYhde18EXZdtff +Zj32l53HOxbgiA4shnIrD4IgrlFk9Fxk3kyzK7W4cskoAzhJaoplYuq+bPlpc5Lz8A6I+Vex1M1 yobZ7tZXevuJ7Fk/sZbXU4qp0ow6I/ro+bJWvNvxyCnvUqQ2mWfnbYsCm/Wa0/heY5MlslvuV91Y 6o5ndTyi06Kq1XGXJtXV16zPdsrnxsXIOgiCI5tsVh+R0CCYRzRFQpvMk4k6tkBVo/Ys8jJVsWL5 qml1nmVebqvX8dQJobqJTU11rpmqAIW0ddUxo1v71jqT66nMs6slEl4zCaiXseT3l8ezOp5qC99M vkmoe62uzrRqZB0EQTCThAgNgqARkTuTnEZd1xoEQRAc+cStbhAEjYhAhgANgiAIZoIQoUEQBEEQ BMHICREaBEEQBEEQjJwQoUEQBEEQBMHICREaBEEQBEEQjJwQoUEQBEEQBMHICREaBEEQBEEQjJxG n1C2LO++++5Ey7sgCIIgCIIgGJTt27e33nzzzea2ndi1a1dqBxcEQRAEQRAEw0D3unPPPbf1/+VN vow76B63AAAAAElFTkSuQmCCUEsDBAoAAAAAAAAAIQBwcs4HbucAAG7nAAAVAAAAd29yZC9tZWRp YS9pbWFnZTEucG5niVBORw0KGgoAAAANSUhEUgAAAqMAAAGjCAYAAAAYdxz1AAAABGdBTUEAANkD QtZPoQAAA/hpQ0NQR2VuZXJpYyBSR0IgUHJvZmlsZQAAeJyNVd1v21QUP4lvXKQWP6Cxjg4Vi69V U1u5GxqtxgZJk6XpQhq5zdgqpMl1bhpT1za2021Vn/YCbwz4A4CyBx6QeEIaDMT2su0BtElTQRXV JKQ9dNpAaJP2gqpwrq9Tu13GuJGvfznndz7v0TVAx1ea45hJGWDe8l01n5GPn5iWO1YhCc9BJ/RA p6Z7TrpcLgIuxoVH1sNfIcHeNwfa6/9zdVappwMknkJsVz19HvFpgJSpO64PIN5G+fAp30Hc8Tzi HS4miFhheJbjLMMzHB8POFPqKGKWi6TXtSriJcT9MzH5bAzzHIK1I08t6hq6zHpRdu2aYdJYuk9Q /881bzZa8Xrx6fLmJo/iu4/VXnfH1BB/rmu5ScQvI77m+BkmfxXxvcZcJY14L0DymZp7pML5yTcW 61PvIN6JuGr4halQvmjNlCa4bXJ5zj6qhpxrujeKPYMXEd+q00KR5yNAlWZzrF+Ie+uNsdC/MO4t TOZafhbroyXuR3Df08bLiHsQf+ja6gTPWVimZl7l/oUrjl8OcxDWLbNU5D6JRL2gxkDu16fGuC05 4OMhclsyXTOOFEL+kmMGs4i5kfNuQ62EnBuam8tzP+Q+tSqhz9SuqpZlvR1EfBiOJTSgYMMM7jpY sAEyqJCHDL4dcFFTAwNMlFDUUpQYiadhDmXteeWAw3HEmA2s15k1RmnP4RHuhBybdBOF7MfnICmS Q2SYjIBM3iRvkcMki9IRcnDTthyLz2Ld2fTzPjTQK+Mdg8y5nkZfFO+se9LQr3/09xZr+5GcaSuf eAfAww60mAPx+q8u/bAr8rFCLrx7s+vqEkw8qb+p26n11Aruq6m1iJH6PbWGv1VIY25mkNE8PkaQ hxfLIF7DZXx80HD/A3l2jLclYs061xNpWCfoB6WHJTjbH0mV35Q/lRXlC+W8cndbl9t2SfhU+Fb4 UfhO+F74GWThknBZ+Em4InwjXIyd1ePnY/Psg3pb1TJNu15TMKWMtFt6ScpKL0ivSMXIn9QtDUlj 0h7U7N48t3i8eC0GnMC91dX2sTivgloDTgUVeEGHLTizbf5Da9JLhkhh29QOs1luMcScmBXTIIt7 xRFxSBxnuJWfuAd1I7jntkyd/pgKaIwVr3MgmDo2q8x6IdB5QH162mcX7ajtnHGN2bov71OU1+U0 fqqoXLD0wX5ZM005UHmySz3qLtDqILDvIL+iH6jB9y2x83ok898GOPQX3lk3Itl0A+BrD6D7tUjW h3fis58BXDigN9yF8M5PJH4B8Gr79/F/XRm8m241mw/wvur4BGDj42bzn+Vmc+NL9L8GcMn8F1kA cXguuI+9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAAJHRFWHRTb2Z0d2FyZQBRdWlja1RpbWUgNy43 LjEgKE1hYyBPUyBYKQBkA4IYAAAAB3RJTUUH3QYRDSYjbPABDgAAIABJREFUeJzsnQWYE1cXhg8u xd2huBQoxYq7u7u7Q9Eft0JxL67FXYpTrBSX4lrcXRYWWTb/fPfuJNmQ3WR3k8wkOS9Pnnhykw0z 3xz5TjiDAjEMwzAMwzCMBoTXegEMwzAMwzCM98JilGEYhmEYhtEMFqMMwzAMwzCMZrAYZRiGYRiG YTSDxSjDMAzDMAyjGSxGGYZhGIZhGM1gMcowDMMwDMNoBotRhmF0z4MHD8jPz0/rZeieN2/e0IcP H7ReBsMwTIhgMcowTCBu3LhB3bp1o/v37xtv27hxIx0+fFizNdWtW5fOnTsX5P2fP3+mcePGUb16 9eh///sfvXz5MszvuWLFCpoxY0aYXmP16tXUqVMn6tq1K82ePZs+ffoU5nWB9u3b08WLF7+5fdiw YbR48WKHvAfDMIyrYDHKMEwgzp49S4sWLaLu3bsbbzt06BCdP38+2OdBBM6cOdPZy7PKyJEj6d9/ /6UBAwZQxowZ6datW2F+zTx58lDRokXD9Brr16+n+PHjU+XKlWnnzp3UtGnTMK8L1KlTh1KkSOGQ 12IYhtGaiFovgGEY+3n7lujuXaStiZ48IXr6VJ5w+c0bohcviN6/l4/DyddXnvz9idTBvw0aEC1f Hvz7QDwdPXpUREMLFiwY6L6PHz/SpEmT6PLly1S4cGFq06YNHThwgHbt2kUxYsQQQrBx48Z05MgR ateunXhOv379qFevXpQgQQLxmki7Q1DhsdOnT1fW/YLKli1LDRs2FI+fOHEi5c6dm+bMmUP9+/c3 vjdS9cOHDxevD9GpcunSJSpdujTlyJFDnFSuX79O06ZNE2tu0aIFFShQgN69e0dz586lmDFj0rFj x6hJkyZiPep77969Wzw+VqxYynf4lrJnzy6eM2HCBBE1rlSpkvIdNhCPRcR406ZNlCRJEvH5IDwt yZIli/hsWO8PP/xg9fNly5aNtm7dShs2bKCoUaNS27ZtxefA91yxYkXKlCmTeN5vv/0mBC0OGDJn zkyxY8emkydPiqgrxKkv/tjB/J3ChQtn9b2DQ/nz0PHjRJEiEUWOLM/NL1uef/cdKd+tPClfoTiZ Xw7PIRCGYSxgMcowGgP9oGgpOn1anl+5QnTtGtHNm/a/Rpw4ckevaD2KF48oVSpShCFRlCjypOgb IRQULUJ58th+vYgRI9Kvv/4qBJZleh6i7qeffqK+fftSz5496evXr1SzZk0hjhInTixS6gkTJqQh Q4YIAXRXUc9jxoyhrFmzCiEF4QSB9fr1aypevLgQR2nSpKGOHTsKAdWyZUtat24drV27lgYPHqx8 tjjifQ2Kmsbr4bq5EAWtWrUSYvL58+ciJQ4hideHcFywYIFIj2NdEMjhFTWE14WgrV27tlhrs2bN hMCEWIPgQ4QVkdbHjx9TlSpVqFGjRpQsWTLq0aOHSL3jsYh6zpo1S6TysVa81r59+4L8Ti9cuEDp 06cXly0/H15r9OjRIrJ87949IV7x/hDDEM7jx49XDjqeCmGNvwmej+8OwhUHDngMShPwuLFjxwb5 d+rQoYPV7zY4FN0qfltfvqAcQp6bX1b+ZIpYD3xd+erFbepBkfnlaNFMwlTR0pQoEVHSpKQIenlu eRm/X4ZhPBsWowzjAhRdQX//TXTsGNHJk0RXrwb9WOiV7NmJqlcnypCBKG1aefr+e6IIEVy3ZtRf IrIG4aKC2szNmzcL4YeoZs6cOUVEFCInefLkighOJdLbAALzxIkTotYTgm7btm0iCrl3714hqhBR RMSuVq1a4vFI8yNaBzEKIM6KFStmfG+IJ0T/sCZLKlSooHy/f9PQoUOFUN2+fbsi6q9QypQphSgF WBseg9eEWIWIjBDwhcZTFDwiwWmVLxrRTwg9tSwBAhmvh+gtnoeoIli6dKmItF5V/pgQ4hDf+H4i I0RoBsQhBDiEMoSxivnnQ1nEwIEDxXeHE77zHTt2CBFcsmRJUQ8LEQnBiwMFFURxS5QoIb5fcFo5 ooFoD+7vZO27DQ7lI4qTozCP3Pv4ED18SPToESnCnwhfubJMcRm34YRIqypMlZ+Y+H+Ak/p/Ardx tJVh3BsWowzjIBD9+esvoj17iA4cQCTs28cgGlSoECmiguinn4hy5ZI7VD2CKCFS0xCHiGQCROqA 2twEwVm+fHmrz69atapIPUOsIbVerVo1kRZHWhii7tWrV4HS2kg5m6eZkUY3J4OizP/55x8R5Yxi JVyGFDhEHATpqFGjhIiDKEOkESASWghfvkL06NGNQlS9b+XKlSIdDhEe3kzdqOvEms2ByEX3uvr6 iGpGQvjZAjVqC8Frjvnnw3uY34/vAl3xiKQiIovoNCKyatRTBe9vHt3EZ8LfzdbfyfK7dSUQl6rA BAHaPkjQi6YKUwjX//4j5YCGaN48IpQGKxqfUqf+VqTiXDlGEO/FMIy+YTHKMKEANXSKzqLNm2XU 0xJFB9GwYUTof0FUyV1TjYhcQuStWbNGRCZR8wnRhMghaiEB0r8AEUGIKhWIT4g8f39/+vHHH0VE cdCgQSKlDxCtm6coCjWaiEigZX2qOUjxQ6ChscqyUeq48gfJly+fcR0QaHh9pK+R2kc00RBQNPsI qsaC+vXrU/78+YVoQw2rOSg9wGdE7SU+M9Lt+E5Q0wmRqkYb8RgIQUvwGEshagleCw1ORYoUEZHY /fv3i3UDfIeIBj958oTy5s0b6HmIyCLCi/eGSEcZAkR7cH8ndwNfHU5BlbbCoACiVD2hvAUZiDt3 iGA4oPz5xHPNT1mz4oDEtZ+DYZigYTHKMDaA8Fy5EnV+snnIHER1hg8nQtDJQie4LYgKmkcGEY2D kER9IoDAq1GjhkhpIzKHJiUIxXLlylH16tVFM9OePXtEBPT9+/fGaCRqGyGwkJIGSBNDLKFBCIIt bty4tDygs8pyDer1qVOnCsGq1miq4HZEXRG9hPBEShspfURHIUq///57IebQcITXsRSNiD6iphU1 ong8wGNwwuMhUFHHiddEVBWfD2UFKDH466+/REQUTUx//PFHsN9lULejRhUiHa+F7xTfZy6EzUmW S6DmE0Le8vno9se6sWaIT3xO9bMF9XcKak3uCg70EAHFyRIcf0CcQpTitHs30eTJsi5b+ZMHEqgo jcF5RN4rMozLCWdQwwUMw4iU4LJlqAeUItScMmWIUN6IWk5EWzwZRNHM09jYTFgKONRAIp1snppG VA+R0OgBYSe8jrn4+/LlyzepbERGcft3ZvlUy/c3v47XB5aCClFTvA7WZA5ugxhDtFAFXfkRLVQH XhefU30fXMZJfR+sAelviGZz0GmPzwcnAUssP4et2/H6+B4s77Ncr+XzfXx8xPurm3Pzv5Xl3ymo 9/Ym8BO6ccMkUtUTbsOxCMqe1RMCyx6k3RlGl7AYZbwa1J/Nnk00Z460RlKBAw+cfnBCPRrDMJ4P mqtOnZJNhuoJVR0IUqviFBkQNBlaqchgGCaUsBhlvIrbt4mmTEGziaw1U6lUiah1a9Q58k6GYRgT MGMwF6gnTkgngCJFZG04TkjxMwwTeliMMh4NIh2zZsFkXHbiqiDd3rGj3JEwDMOEBAyXgGsGuvpx gmDFtqRUKXkeYCfLMIydsBhlPI6DBzEeUjYrqMBSsVs3oho1tFsXwzCeCdy9IEph7YZz1JiqUVMI VHihMgwTNCxGGbfHz0+OLERXu+osBOvFvn2JOnWSowgZxhbwhrXTB55hggUT1NSoKU6wplKFaYkS clIawzAmWIwybgnSZPDxnDbNdFvp0pjiw6l3JuSgeQ0jVM2b2BjGEWAPi8lSauQUmZs0aUyR0+LF +YCZYViMMm4DTKz79ycKsKIUtG1LBPvFFCm0Wxfj/qCeGJ3S5nXFDOMMMHsAjVBq1PTwYeneoUZO MSjDYqIsw3g8LEYZXfPgAVGfPoEFKAQpTjzmj3EU168TYeIpzhnGlXz+THTkiBSm+/YRnTuHARFy ZDCGaQTMmmAYj4bFKKM7UPfZr5/0/lRBSh6ilDfMjDPASNfmza2PdmUYV/LkCdH69URr18oIaoUK UpjiYIlHmDKeCs+VYHQBDolGjZIenyj2hxDt2ROTZeR9gwezEGWcByzAONLO6AFMd+vQQdaXYigH Uvdz5xIlSSJFKUYTY7vIMJ4Ei1FGU2AgnTChtEIZMICofn058QQCdPx4FgiMa2AxyugRdN23aUO0 c6esmcdwjj/+IEqWTI4lxthiGPAzjLvDYpRxOUjDw4YJHaT58skjfsyBhwBdsUJeZxhX8uEDp0AZ fRM3LlGLFkRbt0pfUwzuWLNGNm+ixnTRIpO1HcO4G1wzyriM7duJxo2TRfqFC0sfUGxEGUYrcFB0 +jTR06fyhN9j7NjSoSFiRK1XxzC2efeO6M8/ZY0pBn0ULCjT+RjwET++1qtjGPtgMco4FYzJQ7p9 9Gh5feBAOQkJdaEMozXbtsnUpznt2skRsgzjbiDCj8gphOmOHUR580phiigqyqEYRq+wGGWcwrFj REOGyFqnIkVkpKlMGa1XxTCBwdYvWzaiy5fldZSOwN4JTSQM4874+kpBCmGKrBQaodAYxUNBGD3C NaOMw/D3l12f2JH//LM0coZNCSaOsBBl9AjcG3r0MF2HfRgLUcYTiBZNpuqXLSO6f1+K0O7diTJm JJo4ketLGX3BkVEmzLx8STRiBNHkybKYfuRIoqZN5Y6eYfTOx49ESZPKqTe3b8udOMN4Kpj4hDKU TZuIqlWT0dICBbReFePtcGSUCTVXrsiNGYrkz56VNk3o8mzWjIUo4z7AvxYRozFjWIgyng8anJYs Ibp1i+jHH2XgIGdOopkzZTMUw2gBR0aZEIO0+y+/EJ06RdSqlYyEsh0T4868eSPrRcPz4TnjhcBg H9HSPXuI6tYl6thRClSGcRW86WXsZsMGopQpiYoVkx3IKJCfN4+FKOP+wM6JhSjjraC5CZ6laORL lUpanKHuH96l2M4zjLPhyChjE0z8wJEyNkqoC8Vl3nEzDMN4JmhGhe0ZoqVHjhA1aSJrSzNl0npl jKfCkoIJEsyHR+1n585Ev/9O5OcnL7MQZRiG8VywjUd0FGb6Z84QxYhBVLQoUYkSRKtWEX35ovUK GU+DI6PMN8yeTdS+PVGcOFKQ1qmj9YoYhmEYLUEwAqVaiJZevEjUqRNR166yxIVhwgrHuBgj8+cT RYhA9L//yfoh+NCxEGUYhmEwHhf7AzQ7HTpEdPMmUdq0cqreixdar45xd1iMMkJ4wtIGtaArV0rf UIyQYxiGYRhL0qcnWrhQOqo8f06UIYMcGIEhJwwTGliMejF798oZ8bDygMfcp08cCWUYSx49eqT1 EkLEgwcPyA85VYZxMmnSyLT9+fNyeESWLETduuE3qPXKGHeDxagXgg0H5nHDzgMpFlQNN2+u9aoY vXPgwAGaj1oOM168eEH/Q10HIcK+hn799ddA9w8ZMkTZSX0Ul/v27UudOnUSpx7mMzgDsHW/VuTN m5eeP39O9+/fp/r16zvkNZ89e0ajR4+mxo0bU+fOncV36yjqKkeX586dc9jrMYwtkicnmjpVWkNh kln27LLvABPNGMYeWIx6EQ8fEpUvT5Qjh+yURCQU5vUMYw+JEiWiAQMG0NevX423rV+/3hg5XLdu HY0dO5ZWr15tvH/hwoX04cMHcXn8+PFUpUoVql69OlWtWvWb17d1v9bEjx+fmmG8WBi5ePEi5cmT hyJEiCCEaJkyZWjSpEmBvleGcUcSJyYaN47o+nWiBAmIcucmatGC6MYNrVfG6J2IWi+AcT4QnQg0 IRWPlDxqQuPG1XpVjLuRJUsWSpYsGe3fv59KIaxOMhraq1cv42O6du0qIqUQlJERIrEAwgsiLCis 3b9161basGEDRY0aldq2bascTOWgd+/e0dy5cylmzJh07Ngx8b5Hjhyhdu3aief069dPrCuBskc8 fPiwSF0XLFiQpk+fTrdv36acOXNSz549KVKkSOLz+Pj40M6dO+nHH3+kVq1a0YIFC+jgwYPKQVtl 4zqQ+j5+/DhVqFBBPOfLly90+vRpunbtGrVv315EUAHWunHjRooWLZr4DiA4M2bMaHydLl26iIhx y5YtjbdVw1xdom8+16xZs2jOnDl06NAhIYax5jRp0ojHQeinSpWK1q5dS/nz5xfvEy5gDi8+I6LY 0aNHp4EDB1JsbnlmXAhGRGMyX+/eMmIKA/1y5Ug5mCXKmlXr1TF6hCOjHg5smjB7+59/5Cx5eMSx EPUeTpwgQmYZkQpH0LBhQ+U3tEpcRor+0qVLRmEKflb2OhB0M2bMsPp8CDeckPK2535EXocOHSrE Ht6nbNmy9PjxYyHGBg8eTKdOnaLatWtTwoQJhcDz9/cXQmzMmDG0Da7dhP8Ds8XtEI7FihVTdpIj 6e+//6YlGNCtcPbsWWrevLmIVmLtS5cupXnz5glhu3fvXmPkF+/5ByZABDynSZMmQpyXLl1aiEm8 B24fNWqUWAuEKNL7yZHDDAAC9h/lP6Oa7sda8XlvB+QzLT8XPisYPnw4pU+fXnwP6uP69+8v0vvd unWj5cuXizWrjBs3TqTr8TeCGGUYLcAx0KBBMl2PjBx8StEc+++/Wq+M0R0GxiM5etRgiBvXYAgf 3mDYvl3r1TBasG2bwZAggcFw/LjjXvPevXvKayYwfP782TB37lxDly5djPfVq1fP8OeffxoUcWVI nDix4eXLl4aUKVMaFEEk7g+v/BirVq1qUISboXv37t+8trX7q1SpYti4caPxMYoYNixcuNDw4MED Q9KkSQ1+fn7G+/Lnz6/87o8a5syZI56H9SgC0ZAiRQrDmzdvxGNevXplOHbsmFh3v379xG2TJ082 dOrUyfg6NWvWNKxbt854He/z7Nkz8Z7p0qUzPqdXr17GxyiC0/D06VODIp4NitAUtyli2HhZBY+J EiWK4evXr+J6ixYtxPuFCxdOfBZrnwuXz507J9akvj8elzp1auNjli1bJl4HFCxYUPmbyz/64cOH DcWLF//mu2YYLfjwwWCYNMlgSJbMYKhc2aD8X9R6RYxe4DS9h4EUfL16RHv2EI0aJT1DGe9j+3Yi lDdu3YoGHMe9riLsKFu2bMrva49I0Q8bNuybx2TIkEH5Ddaj3377zZg2VkGkM7g0veX9inikeLB8 CADpZrUGFSlo88eizhQp/atXr4pIIqKVSHVjvbFixRJRzZkzZ4rUuyKqA6XOkRZXUYQrxcHEhwDC BzFyLGJE0+YTJQSIembNmlWk+xHBffLkiXg/cxDBxbrPnz8vSgVQDgDwOQwB80fMP9edO3fE5yhf vrx4D0Rfrb0/1q8cIBivq89XhG+g2xlGS2Ah2L27HC2Knz7cWzBidPBgosKFtV4doyWcpvcgID5R q6Psf4RhPQtR72THDilEt2whypfP8a+PVP3vv/8uUstIy1sDqWaIv7dv34bpvVAfCnEH0JWPWs0C BQpYfSxE2xblQ1+5ckWk23Pnzk2DBg0y1mMuWrRIpPyR3s4UzJDtzJkzi3pRgLT7ixA4eu/atUt8 P0jV7969W9SpWgKngI4dO9LTp0/F9eBsmLYrRxUoH4Cwt2zqgti9cOGCuIz3grhlPIxr14gWL5ad pnCb9xCwj4IgRWMTKlawvYIwvXtX65UxWsGRUQ/g6FFZi4Mg1JEjslic8U4ULURNmkghmj+/c94D tYxoGOqN7gQzEEFUo4hotkHnfffu3UWTEAguIhrU/XiNmjVrKvvhv0TEsmnTppQrVy5Rx2kZdUUE 9P3791SoUCFxHRHQNm3aCBEKIBI7KHtARETxXtnhP0MkXsf8tdAkhCYlRH+TJk0qajXVx6iPs3yO eh3ifOLEiWKtDx8+FLWd+/bto++++874WNUKCyIzbty4oq4ULgJYk+XroqEL9Z+IjOKxaIpSQQQV ohaRYnzviApb/h3MLzM6x8eH6PRpuRH/+295jlSXCqLsKP5OkUK7NToYbBrQx4dtlnK8pRxEEnXu jAZE/L61Xh3jSng2vRvj60vKzplo7Vqi0aPlf2DGe4EQbdRIClFnH5AgJa2KTBVYE1kKSmxeVHFl 7X5bz1dBhBWCzvx+RBTNU9Xqa0B8qe9puU5f5T8N7kf6GilvXMYacbIUbeiwjxEjRqB1qe9p+Rz1 dlg/lStXTghfgGglIsSI8FoDZQj4XObOA5afC+vEWlBqoK4FQrdo0aJ048YN8RpxzboSLb9HW987 oxG3bhEhAn/ypBSeZ87gjx38czAQfvp016xPA9C3iElO0OLjx8uSM8Y7YDHqpmBsZ4MGRAgCrVsn /d0Y72X3bkT+iDZvJgoii824ANgsIZqLEgEY26dOnVpYNVlGccOKuRhl3ABEPY8fl/YWUFpIZ9lT /gGzzh9+INq/X16HNcp//xElS+bU5WoN3F8QIUVCYdo0oly5tF4R42xYjLoZKDOrUEFmczZtQtOG 1ititAbNajgwwe/BSoki42IQjUU9KBqvzNPqjgSbbdSMJkmSxCmvz4QRRD0R7VRP8DKyNdQA0evM meXRpHrCfE2AVMexY/Iy5m1Onuzc9esABInR5ARnMuzn0BMBbc54JixG3QgcIXbtKgu9YXeIInDG u0FPA1JZiIiyEGUYDXj/3lTridOhQ0TPn9t+HrpN8+QxCU/8B44Rw/pjYY9RsaK8jAMciF0vSYe9 eUME0w70ccGzFBHTiNzt4nGwGHUDHjwgKlqU6OZNma0pVkzrFTF6AEIUnagbN8pyDYZhXAAc3M2j nqj1tBX1RG0xopzWop72Ao821JeCnj1lUaUXcfWqDApDh2OqEyY6MZ4Di1Gdg/90+A+IbsOFC2Um h2H27pUR0Q0b2J+PYZyGK6Ke9oLORLUuC8WUEMVemLeGaQT2idDykyYRpU+v9YoYR8BiVKegNrRk SaKLF2XDZZEiWq+I0Qv79hHVrQuDeP5dMIxDCW3U07LW01kD2NHJo87S7NtX+iF5IV++yLJZfPxW rWT63mxuBeOGsBjVIf37S6smzPBdvlx6sTEMQJkGaobhoIDSDYZhQgkmeVlGPZ89s/08TARDyhyi E41FiHq6SgkhFVKzpryM94R4NptQ5m08eSKHu6CkFvtMmOc72LiCcREsRnUEMkJqJgdRrxo1tF0P oy8OHCCqVUv+NliIMkwIUaOesFXCOYSonqKe9oDdNTxrAyZviVbzESO0W49OQCktGpvw55w3D/6+ Wq+ICSksRnUCIl2IhAKk6BMm1HY9jL5AqQYCIvidcAMbw9jAMuoJ48qA8avBgiijZa2n3vK/q1eb 3OBjx5YiO04cTZekF5YulZNTMWp0wAAis1kSjM5hMaoDMD8cXsgQo2vWaL0aRm9AiCIiit9G8eJa r4ZhdMidO9/Wevr5Bf8cRD0zZfo26qn3PC8MOGGEf/myvD50KNGQIZouSU88ekTUvr2cDYBJwDi2 YPQPi1ENefyYKGlSeZmblBhrYFgLyjUw8pWFKMOQjHpCbJrXetoT9cTIVLXWUz3FiuX89ToDNBNg 9i/A50J01F0/i5PAlEL4cjdvTjR8uBxexegXFqMagbqWNm3kZXQGsokvYwn2sdWry4hoiRJar4Zh NMKbop72guJIeBtdvy6vjxwp89JMIODChVrSU6dklJT9mPULi1ENQIYFlk1dukgfUcZ7wYHIy5ff DlOBEEVEdNUqafHFMF6Br++3tZ5ombYFooOWtZ6eHilcskS2jwP4miI6GlYvUw8Fo5JRRwonEnTd R4+u9YoYS1iMupBXr0wuHNjG8vhGBsMMkE66dIkoQwZ5G34biIji9lKltF0fwziVu3cDRz0hRG1F PRHdVDvcYa2E82zZPCfqaS/4nhD9xWg+ANNNeI8yVsH+t3t3uX1FlQN6NRj9wGLURezYQVShgrz8 +TN7hzIy04aABmYvY5+cMiXR4cNE1aoRrVhBVLq01itkGAeiRj1VayWoAhTO2wKd4pa1nugiZ4gW LJCu7yBRIjkrk8N+wbJ5M1Hr1rKeFJ7eqOhgtIfFqAtAbShqRMuUIdq1S+vVMHrh+HGi/PllRPTa Nbl/xrQ/HLXjt8Iwbo1l1BO1nqhLCQ5ENy1rPb0x6mkv+D6xAUFdLZgwQXobMcHy8KHMSn38KLe3 qVNrvSKG22acjLoNxQFsixbaroXRF3/9Jc+RikewCEJ02TIWoowbgr26Za2nPVFPRDjNo56oXeKo p/0gxYYRRPAyAuPGEXXsyK3jNkiWjGjPHqKJE4ly5yaaMsVkTsBoA0dGnQRSr9g+oO4PUS+1HtDb QYkCPJt9fOQJ29Ju3eR9mECFecMqqMX35PsWL5bNsL/+KjeKlSvL3wkeFyGCabCK1uvk+/g+y/uS fLlHrbJK4fn1kKz1jOAffNTTQOHoVaJMFK+SFJ4fchagSbtMUU89fT63uq+DslFNl47o/n1x265K U+hUoa6if6tTJ/m4T5/ksQFHAL/l7Fmi+vWJcuUimjmTj4W0giOjTuDYMVlXD7zVtgkjnlEni1nq cA5A5A+gbAy3YyOKU4IEgZ8HC0EVCDJPvQ+9ByjvAuPHS2eFnTvljGVYJmL/jFKwNGn0+xn4Pi+5 7+NHCn/0NOXae4RSPVBO9w9TrPePTI+jIFD26l9/yksHPhegeykK0N3kBShSwjjUr5+826AIqw+b dPD53P0+jBnClwoPI4X8B8bSPz+0o69foxgfhwN/2BohSIJpohiiwdl8CUaHooKkVy+i7Nlldoo9 v10PR0YdzIwZcpuAA9UbN7RejWtB1FMdvwYRhXpIGLVjI4gNIGNi3z6TZRMOVswbiKNFIypbVjYx 4TLDuJR7977tcLen1jNjxsC1nvCw41pP14AyibRp5fghgB0R0vUWQJRi2h+mE6GJB6B2/cEDmZnx 9sZaBANatpRCvXdvrVfjXbAYdSAYF4wUNH7IqCMN0XW8AAAgAElEQVT3Fi5ckJ8XjVqqXRUm1nGX YtAMHCjT8wDfE+qW0D2PE8R7lCjBP59hHAJEjPk0I9g5oLvDFsgBwxtHtVbCf3yej64tKHyEdxFI lUpGQ+xQl8juo17y6lWZ1oc7lDfPdIeeh7UepiPCytXT7Wr1AotRB4Ef7Lt3UozCWNcbwFE2jq4x yhTR4D59vLMkITQ0bSrLFCA+MV0Jnt0M43QQ9VStldSoJ1IawWEt6okOdz7a1Beogfr+e9OQgNmz idq2tfvpqF8fM0Y2VDZo4KQ1ugn4L4GgEkrKNm6UQX7GubAYdQAY/IFxY+fOyZoTT0eNemKqBcza e/Tg5k2G0R3oWjHvcA9J1NPS19ODjpZ8lKPotopIi2E2rWjOnDnGy20tBJze7/NVROjIkSMpNbqT kKJC8SOAMEX3bAgjBNC0ankQbI8aNgzR0z0KfH74kWJSojd/D66A41hhAJFBREEhRDHS0YO210GC +iLYYqRIQVSxojRoZxhGByDfalnraU/UExYOlrWeHhz1XL9+Pd25c4eaqaM0LciD6EIQ6PG+LVu2 0OzZs2nUqFHSwgWTmDCUHR2Sf/wRYk9B8zp1jM6EIIM1ITz1vQ0IUPQ7IG2PYzm4nnhzCYMz4cho KMF2HxNzACbpePC22wgOyjGxApk+bqxhGA1B1NOy1hNdKLaIGVPWenpo1JMhKUbhPQrSpye6cuXb lnw7Qd8aNO6sWTIL5q0/lbdvZWkV6knXrjXt+xnHwWI0FMCX7Mcf5WVv+fbQnIT9HVLz2L4xDONC zKOeqPlEOoajnow10LwAOxOk6wC6cDBuKAzs3Sv1bLFiYV+eO4OZApMmyTpSnm3vWFiMhhD8p0SB N7qd0YjqDaBGFE4BixYRffed1qthGA/HMuqJU4ChebAg6mlZ6xkvnvPX62bcvn1bnKdRTXzdnOfP nyv68x19jxpRlZEjiQYNkpcxXhVhzTAehJhb93kz8IOG+wAM8r2lWdkVsBgNAWvWENWtK4ML589r vRrXgV8I2wUyjJNAet2y1hOC1BbocFetlXBC9yRHPW0ybNgw5QDbX5x7AvPnz6fDhw+LcyNwt4fY fv1aXkfhpwNa5GG4DwcV/NwwqMNbwf4fPRPt2kmbPibscAOTneD/Of4TwowcR0aeDqKh2H6hQZOn UTCMg0B4ybLW056oJzq/zWs9IULjx3f+ehn3BDMtMUNUFdwwNcbMyzBGFaJHJ5o2jahwYflT7tnT AWt1Q3Dcd/KkHBQAf9Z589gbOqywGLUD/OeDvQOioqtWab0a14BRlCg9QjSYYZhQEtqop2WtJ0c9 mZACMYoCR3TfYCYzOm8ckFfGMdCePXI4BxqaMLHIG0mcmOjvv2U5LqbpoZ/Ccrw1Yz8sRm0AmzYI 0caNpUuGNwCrun//JfrnH07PM0yIQG3erl0m8QnrCVuoUU/zlDtHPZ0Gais9qTotUaJElBajQC2B UkQuXR31hjrS2rUdslFPnlyOFH3xQvZOeKvPND43Bt0MHizLtbdsYYP80MI1o8GAKCgyG8BbviWM 9oTohjMIT/djGDvBBgIp0eHDbW8szKOeEKCIeobSeodxPBOUo/HSpUtTzpw5jbftUg4w3r59q2i5 2sq28X9UpEgRqoiiQcLxxj3atGkTde7cma5fv05Tpkwxit2flb9vE4tOdvUx4RRRmFxRdYULFxYn c7Zv306HDh1SdOSvYfswUIuoHYUpNtiwQZpmMg5n6VJZtrB+vYwaMyGD8z5BMH26FKL4UXmLEAU4 qsNIOBaiDGMnSINi+gPEqOXGAlFPzHuFQS/CJs+eyXTL4sXSoBwecSxEdcX79+9pFow1zRg/fjxF R8GkwuTJk6l169b0MsA66eHDh7QRXj+EKXzn6N9//1X0XnVxyp079zevj8ecPXtW3J8wYUIhYnup U5MIHvUtaNKkSbQWafWwggh7x46m6yNGhP01LfjrL1kB4O0ge4qyNmh9jBFlQoiB+YYFC7BHMRia NtV6JQzD6JorVwyGTJnkBkM9FS9uMPz+u8Fw5ozB4Oen9QoZM3x9fQ2//PKL4datW0E+5vr164ZE iRIpfzr5t3v+/LkhSZIkhs+fP4vrMWPGNPTo0UOcwNGjRw2lSpUSlxUBaahRo0awa7B8zNu3b8X7 ncHvxWAQ55cvXzZkzJjRrs90/vx5w5EjR4J+wJMnBkP06Kbf55Ytdr2uvShfqSF9eoNh82aHvqzb gj+F8uc0rF6t9UrcC46MWoADXBRk4+gGwQtvAaVuGHv2/r3WK2EYN2HbNqL8+WU7rQoiXAgVdejA UU+dAT9OReDRjh07aOHChUE+Ln369GLO+16YShMy2xtEFDNSpEjiOmyhBgwYQMuWLaObN29+83wf Hx+6du2aOL1WrZWCQRG3VKZMGTqCGmPCz+bHEH0uRGVLlixJXzEK0BqY44kovIqDo6Oom4TbDHor bM1h8AZQeYMGL/SPmbttMcHDYtSMAweIatSQthUorfEmUOeOjQmb2jOMHaCWpUoV6ecIkMJdsUKO aOGud13xKcC9ALWdy5cvp7qwRbFBw4YNaVWAdcqaNWvEdXPix49PvXv3FvWjlly4cIH69OkjTnsU VXLmzBkhXHGCILYGSgA+hnKKStKkSSlGjBiizjRIlLUau4yOH3d4HrloUbnfnDvXoS/rtqAMHJ32 0P1oCGZsw1vNAC5fJipeXNZ640fkTaBr/tYt1CppvRKG0Tlw/cY4sn79pBkvgBkv/hOp3Y6MLkAt Z7Nmzeh+gI9rrly5vmkUCor6yt8STUmPHz+mGzduWH1eF+UI/sSJE3Ty5MlA3floWkK0Eic0PGE6 0pMnT8TpC4a9WwF1pJkwKSmUIOI7P7gwXJIkRG3bmq47oXYUja+dOjn8Zd2WdOnkZgEepOowLCZo 2NpJQdlGUNas8jJEmbcB321krTijyDDBgDGSaFRShIMRHMGia4ENBnUD0uijRo0STUDt27f/xvbI HmunJIp4Q7q8Y8eOQlCGs2KHFCVKFNHtjuajzJkzB/laRYsWFSdrYB3Tp09XjnE+UFlMVAkFsHZC mh+fG81X3wWV3urTh2j2bOlzi40+csmlS4fqPRn7gAUWqj2SJZPXnXAM4DF4vRjFgSoOGnFQeuWK 1qvRBmRwGIYJBuxRkN6FVY4KupSnTFG2ol6/GdUVu3fvFpHG06dPi9pPS5o2bWrX6zRq1Ihatmwp uuPNiWj290YEdc6cORQtWjRxPXz48OIUHBCLWGOWLFnEY/PkyUM7d+40vi6E5YMHD+jOnTsimjtk yBBRsxoUVapUESebQBlhmsnvv8vrUEYOFqMwGEA00J7leAtJk8rvBU5uqOqZOlXrFekTr/cZxQEv DlpRL8owDPMNGMH2yy9Efn7yOub+zZzJdS06AulypLljxYolLJfixYvnkNdFWl1tXFJBo1AEizQS dqNq9NTa/brh7l10aMkoDNi/n6hYMYe+RbZsMgBrZ0WE1wAhCu0PvcF1pN/i1TWjaubFG2bNBwUm R6BMgWEYC9BQ0ry57OxThSjybThyZSGqC9S6UEQOowQMB7clRG/fvi1O9mApRIE1oWmexne1EEVT 1K2A+rIePXqIGtcgSZVK/qZVMKTBwSD46i3TCkNC7NiI2hPt2ydLzpnAeK0YVd0z7tzx3lFmZ87I BmDM2GUYxgzMlEcIw9zfDZ4tJ09KOydGc9Clnj17dkqRIgVdvXrVKEZtsVj5mwZn7eRuoNFqJEZ9 ErbliWns2LHBPwEOAGqpAcpPkFd3IKgowJx27865WgfDZFCqC+MDbmoKjFeK0QEDiM6elRFRHCh6 K3/+Ka2sGIYxAzvnPHmQ+zXdhnAPIqIoALOJga48PE4+H185bYkM0atXr0RdKJqIYG3EkIgSr169 mnx9fYN+0PffE5mPKHVwdBT9YkhDP33q0Jf1GBC4xzEAxoY6ITDttnidGN21i2jUKHlUEsrmRY8B FlZFimi9CobREfBhKVmS6PFjeR0RJHQc4PbIkW0+HQK077IK1GFufmoyPROdvbPfuev1IlAXWshs 6DeM3q01KHkz8ByFDRW8UYMF42nVcgLsFI8dc+g6GjXijFtwYEorynWXLycaPVrr1egDr2oDxZFa uXJy/rq3H5EghdK3r8Nr1xnGPUFDB2pDzWeSJ0wobZvs/E9y+9lFGriyGj14+Z+4/vr9M+q5pAx1 LDeBaubr6oxVewWYaASLJRjIj3bAnhuenOYG80WUI3Lz6//8849y3BE5zPf5+n5WtrMRKHz4Lw57 TWv3pUqVylgzCtavX2/bQB9NTDDyV4s7sUPcujX45zAOBZsXJGEQEIIZQ/fuWq9IW7yqm16tMfee T8wwjE1wlFqrFtGhQ6bbfvpJjmGzs47n4OW19NvGFuT72cfq/WVzNqFeVeZQpAheWqAeBiZPniya dDDtKEgPzRBi3v1+6tQp4dGpAqslR9y3YUM8MRehRo3nDnvNoO6z1vVvE4yxhcG2+pqoh86dO2Sv EQRILIwfL09M8MDgAOXocNwKxsHL4/EaMaoKUfy/s+JfzDCMN4IdMAqnA6b0CBAxQlo+wDsyeAy0 aP9QWnJwhNFIPVrkGNS5/GTaenoeXbp/1PjITMly04h6GyhhrJQO/hCeB6YXqd6ab9++FZZN7gRK NlGaiZ/Qf/+5fkIs6mkhVOOgYyY48FtHFyvAQAfle3cESDSgjPf9e7bhtQc0E6NscMsWKUy9Ea+o Ge3cWZ5fusRCVAXF05Mmab0KhtGQpUtljkwVoqihw8z5ZcvsEqK+n9/RwJXVafGB4UYhmjxeOvq9 9VGqmKsVTWlxgCr91Nr4+KsPT1HbObnp7B02NQ4K1IUWLFhQTCVScTchCmBNC8s8OEg5SN+FCFhe ZcuWLVAk1Sro5lV3ips3y85eB4AgbcqU8vMztsmVS1ZM4PgrOGcuT8bjxeiRI0QzZsjuvixZtF6N fsCR2Lt3Wq+CYTTg61eiXr1kR7FaWxc3LtG2bXJkoh3cf3mNOszLT/9c3Wy8LV/6cjSrzQlKkzCb uB4xfGTqVWUu9aj0O0WMIFOoso60NG04Ps2xn8kDWLFiBVWrVo3atm0r6kPdlVevAjel4PjG1UCI wvJqpy0TbTjUo0QF4IDKgfMqURPJHfX2U7480a+/yr6W58+1Xo3r8Wgxin2OcpAt/lNggApj4vVr HqfNeCGYy4etvvkIFOyQYeNkp73GsRvbqMPcfHTn2WXjbXUL/EK/NdpGMaLG/ebxVfN0oEnN9lG8 GEnE9a/+fjR1e1f6bWMz+vLVRqOJh/MJc9IDQEr5+vXr1NzclN0NgRDF9lXl+HFtJvxhlOmCBQts PxDWMmp0FCmzixcd8v4TJ8pSBcZ+4CCHyonKlWWphzfh0WJUrVXho7NvwQ/dDqcahvEczp8nyptX uk6roF706FGidOnseomVh8dS/+VVyOfjG3E9SqRoNKjWcupQdgKFC2Zz+kPKQjSn3SnKmsJkmL/z 7BLqsqAwPXt7L3Sfx81BJLRixYrG6xUqVHBYg5JWoFYSxzsINmJYF45zMBZTiyl/8ByF1dN7FG4G R44csl4UODA6ilnsdtnyMoHA158xI1GDBqbeMm/AYxuYunQhQtkRysGSJ9d6NfoDZXGZMzuseZJh 9M3atXIMorpjRiQIs3CHDLGrkPyT3wcas7EF7bu42nhbkjipRUNS+iS57F7Gl6+faMq2TrT19Hzj bXG+S0hD66yhnKm9w2ft5s2b1KhRI9EBPnfuXMqVy/7vz52oU4eobl15rntOnzbtDNBthegodhCM JiCri3R9lSpE3bppvRrX4JGR0cuXpRBFuoSFqHXgc3zhgtarYBgng9ACRCcUgSpEY8aUXSVDh9ol RB+/vkWd5hUIJER/TFNc1IeGRIiCSBGiUK8q87y6jnTq1KnUvn170azkqUIUoBzZbUZNw8oMuWGA /zMB40XDwuTJDp806jWgl3LlSllNpEUDnBZ4pBiFdRro10/bdTAMoyFv38r0o3naMUMGOW2malW7 XuLM7b3Ufm5e+u/JOeNt1fN2pAlNd1Ps6AlDvbSg6kjHbGrucXWkGE2JefAqmJ2OFHI4D7c20YsY HawcjN2FmaUtzIelQwldvx6m98V/M3velrEOejpQwtumjbSE9XQ8Toxmk42sXlVrwTCMBdh658tH 9OefptsqVJDdJHbaaiBS2fuPsvTmwwtxPXLEKNS32gLqVnEGhQ8XdvNE1JHObnsyUB3pjn8XU5cF RTymjhR1oZh4dNvM4yeylxSrozcrShStV4E1RLFvchX+v6C5DyBPjLnZjKbkySPdGHBM7WN9nobH 4FFi9OBB6SUKhxYPP+gOM/nzc6cj46FgrCF+4ObhBFg2QZjaMgEnEpFJdLojUvnV/6u4LUHMZDS5 +QEq/2MLhy41Qczk4nUr5mppvO3qw5PUbk4eOnf3oEPfy9X07NmTJkyYQKtWraIhqM31MvQSGYU7 wcqVK22PCAXm0VH48JqNGWW0oaWyaShenKhpU61X4lw8qoEJAhQdjA8eaL0ShmE0ARGggQNNqZHo 0Ylgb1Ovnl1Pf/7uPg1cWUMIQpVsKQvQ8LrrKF4M57YGbzr5O03f0Z38vn4R1yOEj0idy0+i6nk7 O/V9HcnLly8pXrx44vJff/1FJUuW9Ph0fFCgSR16DudaA6eCxo0bi8Yxm5QujT+evNy6NdHcuaF6 TyQhsD9OkSJUT2fM+PxZzueoWZOob1+tV+McPEaMohQMkws849MwDBMi0JyEbnl0zaukSSOr/3Pm tOslLtw7RENW16aXPk+Mt2GCUvdKM4SBvSuwtobyPzanXyrP1PVce9SFjhs3TtkG36AlS5ZovRxd kCmTHGqEc63x8fGhN2/eUHJ7OnqRYiwW4OyAUUrYsaZK5dwFMjZBkA1pe0xqwvGCp+ERYhQ2gfA0 w1xXtSGQCR4ctSZOTJQ6tdYrYZgwglQi5uidMzUZUYkSRKtX2z3ZYevpuTR5WydjVBKd7p3KTbQr KokdPaYGxcAw7gDmzJljvIz7zLF13/N3D2jwqpp0+cFx4+1RPiegpf3OUIKYKUL1mpb3IWU7c+ZM h/h6YspP69atqVChQmKMZwKepiHAthW6zi23scgLq0797duT8mPRdDmMBL8nmOJjUxeQgPAYwl6F rwMgRFEozkLUfjAitWRJGCNrvRKGCQNIJyIF/+KF6TaYDGP8S0Tbmzc//880dVsX2nLKJNjifpeI htZdQzlSFbVrCf/++6/yVhEpD8IWVgjq9qDuQx3plBYHqf2EknTT97C47VPk59R2dm7jukL6mpb3 /fbbb8Ljs3v37kE+1l7QpIS6UMyUZ0zopWZUBdOuUDJhVwMZ7NBKlZKXUeaC0pcQ+iTC2gkzJpRj FMZBFC0qxSiqJ9Bp70m4fWS0fn1SNoTKTsVPenMx9gERymKUcWuwt8OM+a+yyUgckc6aJdP1dvDq /RMasroWnb9rMkPMlCy3MLJPGCulExYccjaemEEzdvawiNiijrSTZmt68OCBiIZi3CQTNOiVg4mA HT1zLuH8+fNi4tWdO3cofHg7epcxOko1Cu3cmWhayHxwMUEIDmo4ZxwH6kdhfNC1q2xu8hTcupse 838hRMeOZSHKMF4DQk5oLe3RwyRE0SmBHJadQvTqwxPUbk7uQEK0TI5GNLXF37oRogCic2LTvyhe jMTiOkTplG2dacymFmKakytBZG3YsGGUI0cOih8/vkvf2x3RW2Q0e/bslDhxYtpjPg43OBAdVZk3 j+jRI+csjAkRCGwvXy4NQm7e1Ho1jsOtxWjcuPK8d29t1+GOsLUT45Zgvi/aSlHFr4L08KlTMlxg BzvPLqauC4vSs7fSdiNC+AjUvsxY6l9jKUWOGC3ES9q7d6/9O/hQkD1VEeFHmiW56fPt+HeR8hmK iO7/0ICGI5xCQqlSpejKlSt05swZqqbOMmesgnwjfEb1JEYBotkLkHa3h7Jl5Y4CQFmPG+e8hTEh AoN9cKwAcwT1eNzdcVsxinGWAL6iTMjp2FHWnzCM23DokGwnPWmyXRLFU/v2ESVJYvPp/gY/+n3X L/Tbxub02U96LsaKFo/GNt5B9QqG/oj277//FidngsYl1JFWyGXyOb3y4ISoIz1/N+TvPWLECJo0 aZLNx92/bxK7GOOJ+tBU3FltEwhRuIrpDXiOor75vToa1xbmvqOzZxM9fWr3eyFxgeNGxjmgNB6T jT1lNoHbitFy5YiSJrV7mArDMO4MOsJR5PwkwPIIljPTp0sPRDsaMt76vqDef5SjNUdMAixt4uw0 q+0J+ul79/BJwVz7PlUxAWq6ca79q/dP6ZclpURtqSNBXSh8KZctW2a87SfML2fsAj9TPZoKRFcU 8tKlS+13UahUiSh3bnn5wwei8ePtfi8kKthj1HnAvhcuamhGNj8+d1fcUoyisQ/cuaPtOtwZWDvx 98foni9fiNq1k6cvsomHEiWSXfSd7Gvi+e/JWTHR6PStvcbbimWtRTNaHaGkcdI6Y9VORa0jRdc/ UOtIx25u6ZA6UnTaoy40bdq09Msvv4T59byRZ8+IEibUehUOwjw6+vvvRM+fa7cWJhBICCHBgfn1 7j4C3S3F6K+/SnuDSJG0Xon7gqOp/fu1XgXDBAPCS/ALNfPJFFEahAHszP/tu7iKOs8vSI9f3xbX w4cLTy1LDKehddZS1Ehh99gEqKUsU6aMQ17LXlBHOqfdKcqcPK/xtu1nFlK3hUWFT6ktEBmLHkQe +fr166IudPjw4co2ljeyoQHZ7ESJtF5F0GBM6yN7G5LQEq8OjkB6H7ZpdgCzi3/+sf04JmzArQCO De5uBet2YrRKFXlulj1iGMbTOHFC1oea781QrY/azJS2u90N5E/z9w6g4Wvr08cvH8Rt30WJRb82 2ERNig6y8eyQUbhwYXEKCuz4z549G+i2Xbt20dqAaVH/+9//aNu2bcb77t27J8zjAYRh586dqVOn TuL0h1njFupI0f2PCU0qMMpvOCEj7TlhNonKCgMGDDB6jB47dkzMLleZP38+14WGEb2LUdSM/oqo jj0gH2weHcVv89Urm09TflZ0924oF8iECASXhg6VEXl3xa3EKLxE//wTaSStV8IwjNNAIRS669Tm Gfi2oZMXg76j2e52f//pNf1veWVa+repsj9Vgkw0s81x+jmD6ydjYMc/C/6nZowfP94YmZw8ebKY YIS57uDhw4e0EWNMCZNWzglT/erVq4tTbrV+LwDUkXYvP4s+nE2mKHA5A/5LuA80als9Mes+ONCc hLrQWrVqUU47R6Yy9qH3ND0amdCMBrsuu8BQ9GzZ5OV372RumNEN6K6Hqx3sntwVtxKjqstE377a rsMTYGsnRnfAowQ1ipjE8FF2u4uZd9u3S3N7O7j7/DK1n5uPjl3fbrytQMbKQoimjO+cIeG2rJ0a NmxI69evVz6e9GB58eKFMCBXU/tIhdevX59Gjhxp9fmJEiUSj8UpK/Y6FmCSUsLPuWhKy/3GOlJE hidv7UTjNreyWkf6+vVrypQpk6gLvXr1KmXhTlCHovfIKCLfOLDZsGGDfU+wjI5OnUr05o1zFseE iiFDiHbvJjp8WOuVhA63EaPoXTh92v3rIvQCWzsxugLjPGGRYR5x+eEHma63sx7zn6ubqMO8/HT/ xXXjbY2K/E+k5qNHjuXoFYtIJma827J2Sp8+PaVOnVqIVgABgCinWo/p7+8v0uboXL9pxcXax8eH rl27Jk4QkZYcOXKEKlSoIMaEoo40UzLTONBtZxZYrSNFkxIsfmLEiOGQ+fRMYPQuRgFKM3LlymX/ E+rUIcqcWV6GEIUgDQa2dnItyn9lmjBB7tvd0XvUbcTozz/L8/bttV0HwzAO5tw5OcQaHfIqSAsq IovS2tPtbqA/Do6gQatq0IdP78Qt0SJ/R8PqrqXWJUdROCds5g4fPixS23aNVSQZHcX8drBmzRpx 3RxMNOrdu7eoH7XkwoUL1KdPH3FCBBbNRRCuOD1//lycYseOLR6LOtJpLb+tI8W0qQv3Dhlvw7o7 dOhAixcvplGeYlSoI/Sepgfx4sUT0XG7wW9dOWgygg6ld++CfDhbO7meevWwLZGmB+6G24hRREVt HIi5nLdv3xrrvJwBvP78UCjrBNjaidEFijATE5Ru3ZLXkQ4cPpwIzT041LeB7+d3igitSQv2DSYD xt4oJIubVtg2Fc1SyylLHjNmDNWoUYNmz55Nke3wOAVIw2/atIkeP35MN27csNrw1KVLFzpx4gSd PHnS+FnAz8qROGpIcapdu7ay/39HT548EacvX75QmjRpxGQklUgRolLfagupa4WpFCF8RHHbS58n 1GNxSdp80pRaihUrlojoHjpkEqmMY3CHyGioQOt2hgzyMvZ9AY12jH7AnwS9aR8+aL2SkOEWYrR6 dXmOiQOOAjsEdKniBAuTO6FQZgsXLqRxDhiRFlTHbN26dUUDgzNgaydGU2CKhyiL8hsndRqMIo4U xSZr08KFs/kSD1/doI7zfqZDVzYab8udtrQwsv8+UXanLBud7ps3b6ZTp05RxYoVxW32WDslSZKE fvzxR+rYsaMQlOGsfL4oUaKIDmdbkcqiRYsK/0+ckiZNSk2bNqU5c+aI2k+AMZ87duygGvm60MRm gf1IJ23tKOpIo30XRTRQITq2bt0642s76+DX23CHyChAc1wGVVzaA5oJ+/c3XYfNUxDTnNjaSRtQ /o35IFOmaL2SkOEWYhT7J0d7L8NqBUIPnaRRo0alfPnyOTXKCVA79syK94KtjlmG8ShQbwbvQnPR lTGj9IJRvdtscOK/ndRuTl66/cw0D7hOgR5itGfMqPEcvWLR4Q6wrTh48CClMMs/2rJ2UmnUqJGI buLcHNRuqiCCmlH5LqIFuAYgnW6rFKBQoWBGT1MAACAASURBVEKi+amksgdCI9IPP/wgRDNAHSnm 2lvWkT5Osp0at6ojrkczcyiAGP7gbiEVHeIukVEcIKFu+C/zEhlbNG5s6n6FAX4QOWG2dtIONDPh OCGYKgrdoXsxCu8sgMJcR5MgQQIqUaKEqMVCg8HRo0dFahx1Ww0aNBBF/l/UqS8Ky5cvF1GIwYMH fzPbF9EFWGUARDq7du1Kbdu2Fc0FAHOgnypbKLy2+jhzbHXM3rp1i3r27CksObAOABGLVKFKv379 RP0YQE0batMYRlcgnYxisq1bTbchwoi6EbU5wgarj4ynfssqks9H2cwTJVI0GlBzKXUsO5HCh4vg 0OWiuQhCb/t22Z2fMGFCihAhdO+B/7uw0sF0I3PQXa+CiOm+fftoy5Yt4jrKAdRa0+DAtgbbLmzD kPVpg5EsASSMlTKgjrSZ8bZL94+JqVQX7gUOXaFBCtsga41SjH1AAOD4wg4XMl3QsmVLWrBggf1P wIczj45iRKivr+MXxoQalAJjs+pODly6F6PDhuHI37nvgdpPRBIQuTx9+jQVK1ZM7IBQT7UEnodE Yp4vzKvxHxfCFdYsKkjbIbWmbsQrVaokUuz16tUT5/DzK1u2rOharVatmqgBsyS4jllcL168OBUs WFDUlU2ZMkVsPLBjHKIcAmGHefv2bVHLpppnQ6T6BzMfjK2dGJcDk2D88JTfuBHlAIogvAIacILj k98HGrGuAc3c1Zv8DfK3nSh2SmH8Xjp7IxvPDjmvXr0S/293794tDk6Dwpa1kznWJhpZE7fmafyQ iF80MlkrAZB1pIuoS4UpZnWkj6nH4hKB6khx0Jxf+RsV4TboUOMuKXqVFi1aiH2MZYAlWGC/pg5G QBjYwkeX0Z7Bg2Wq3l0cuHQtRtV6k127nPP6iB6WL1+efvrpJ9HJivRYlSpVhFhEtCJdunQiygAQ ZYQQhChs1aqVUVBiesnAgQNp586dItKKCErKlCnFf27UbsHPDaI2W7ZsxnKA762oQMuOWXMQJUEa ECUFSOEjuorpLagXQ/MCmh6ww8TaIUbR/IAdJOxegoKtnRiXgWYcVNQjNa8c+Alg+I6I3+jRskvX Bk/e3KHO8wvR3gumSUE5UhcRKeiMSZ1T1oJIJlLgiFQGNToT2LJ20hM183WlCU33UESDDNupdaTj t7Q2+pFOnDhRiO/Pnz9ruVS3xV1S9CpI02/dujVkFl84qDJ3fkDvhOoNHABbO2mLIl9Ev40zssrO IKLth2iHWoYVzH4gTKBeBhFGNBeodVloHpo5cyZVrlxZREshUIEqTi1JliyZSMsj+gkxChGKjbha s4UJJ4XsCO1C3MIY2xqI0MD6RQXRD9+AtEhVZQePDQmaF9CIhcgrBDLEL7plGUZTEG1BFMWsSUY5 gpKF4Bbp6qA4e2c/DVldh958eG68rWqe9gHd4o6dnY5sgrotQDmOJ9Zv50xdjHJ+aUnXo2ygtwZZ C7v19Hy69fQCDa+3nuLHSEb9+/c3mvQj64KDXsY+3E2MhpoWLYgwqOHBAyLMuZ83j6hzZ+PdqMZh tAW9oNiEYfJvPMeX0jsU3UZG1aZOO7NfoQLRDohJ8waBRYsW0dChQ8XG2NyDDeJuf0D7+Zs3b+hu QGU2oqDzlP+EderUEUIU3oNId6BmCz5+7du3F48BsIF5ZcdMX0vwmnhvNVKBTlmk7AHEJyKnsHaB uMbOc5DyC8TtwcHWTozTgYF7gQKBhSjaPE+etFuIbjwxnXouKWMUopEiRKbeVedSj0ozHS5EHyk7 VNReqniiEFWJbIhFeSO2pXI5mxpvQx1p29m5jXWkankA6t9xgMvYh7ul6VUwyUvtObCLKFECj0Mc M4aIo+m6AseQtWu7x7Ag3YpRHHSBUqWc8/pBdanCjBoiEilupM7VxyA1jkYCRDAR6UTnPWqzcCpX rpx4HnZkEInoaoWARHQVaXm1ExcCEWUBlmP/glqLejtqWFEekD17dvH6akofQCRD/KLUAOA9kaJH xDQ42NqJcSo4ioSRvVltNXXrRrRzp3RltgFSxmM2taAp27rQV395ZBo/ZlKa3Hw/VczV2uHLRZkN DubyqzOHQ4C5tRNcOvIqn1s9mYvb0NyHwLKjXxOgMTNj+izUr/piq3WkW06ZagAxHQrbFXvrYr0d d42MIrs3YsSIkD0JjXJJk6ovQGTWCMXWTvoAwWoYHujdtS2cwdxdWUegBh/dYOZNt44GaShrzQFI gUMEwvfPPG0HIELjxIkjbsNXh5N6Pzz6VJsWbOwRQUXq3hzchhody/e1thbL2xAZxeta1vbgcViD 2riAx1hrlDAHmVMEqZo1C/ZhDBNy4CmCgyV1Jl3UqLLBwc4f2wufhzRoZQ0xOUgla4r8xhSyo4Fn KA4U4XIR1sYdNCJevnzZeD1mzJiUOcAlIDT3YULq7Nk+9OyZ414T4ADWfDty9s4BGrqmDr1+b7Ke q/RTa+peaQZFDB9ZERX/iM7+//77T7wOEzS9ehElTy5rJt0JlGPkyZNHBE/sHeYgQMu26r2YOjXs ZERNKXr+EBMJpvePcRHFiklRWqeO1isJGl2KUfyWUaoJiww7hrAwoYDFKONw0MCASMnSpabbsFfe sEFGSe3g4v3DNHhVLRGhU6mQq4VIy0eKEMWhy1UP2hAVRUre8sBRaxBURjUDdiCrVzv//Z6+vat8 9zXp6sNTxtuypvhZOQhYJw4CEGmFgIXrCBM0tWrBL1bfO/6gQJQfpWV1QrJ4+NKiKRchYTB3LlHr 1ixGdQQG2mGC5cGDWq8kaHSZpq9USZ6zEHUebO3EOBQ07KHj0FyIonEP9aF2CtE/T8+hHotKGIUo UsfdKk6jPlUXOFyIounv0iVpmI8yG70JUbB4sTyHXfCOHc5/v0SxUtG0loeobM4mxtsu3T8q6khx kIDSI9jaAdV3lTHx/tMbuv3sAp2/9JYefF5HS/8eSZcfHNV6WSECFobW3F6CBR3GCAerwCFD7zlh L6NGDaL//oM3udYrCRpdRkaRbUbU310sCRjGq4GtEark1cgIQH3itGno2rP59K/+X2j6ju608YRp kkuc7xLS0DqrKWfq4g5dKqKhqP9evXq1SGGHyM7GhaDCAUOeHgcEiKEBL1xw3QH6umNTaOauXsZ6 XTSOdakwlarkbieu9+3bV5QkYWKTvcAtBKb/KqVLlxZeyQATgJ6a/X70dl+M2FHp2dv7tPPAZrr/ 7D967/eCvosXkXz9X9PTt/fo6Zt75PvZRzz+4K8+VLBnMooYVdqY5UtfnlqUGEaZk3lwezmKm9Et ozZALVxIx7M2p2TJ5O+Y0R4MvENPKUwP9IjurJ1On5bnDhj5zjCMs0EtaNeuUHnyOmqVkQ9q396u p7/+8JSGrK5N5+6YfDozJM1FI+ptoMSxUzt8ufDrxbQzjN/VqxAFiIQ+NlUqCOeLAQNcN2+6Vv5u lD5JThq6pq6oI/3y9TNN/LM9XXt0irpVnC4cO9BUiRr6mXa06h4/flz4JGPcqQqaLlUBiJpU8/pW V95nCO9HydLHoZg+kRXBeY9WHVpCj1/fIf9IvvQ14gdactOPPvpZMYR/8e1Nn30SUfiIH41CVHz2 GzvEqUDGykKUZkjyk83vy+3A/yVEkNTJTIryyXe5iZxlz+gCxAdQYTN2rD5tnnQXGcUkTGwj9LUq zwPWTokTy4gLw4QYWLigIh71YSpoIYaNkx1z2sG1Rydp4MoaIuKkUjp7Q+pddR5FjuicWYqYMNTD DTpL4CaCTmRoewSdUHeHgTft2tk1I8BhoI4UzWTXHp023qbWkUahWGLQBr5TOBHokS9fP4rfFyKX OH/y5q44h+hUb3/r+9Ih7xUtcgwyPK9E/24cSj0mjaVPXz7QgUtr6av/10CPK5y5GjUvPozSJc7p kPd1NKgNhvNCiO280OSBHYpqX4iSnUaOn4zGhJ6WLWUdOnxH9YbuxChS9AiquIMvljvDDUxMqEHI Dl0ahw+bbsuTRzYq2ZmT23N+KY3b3IY++8mpLeHDhac2pUdT/YJ9HLpUOFBs2rQpZA0ZOgIVEBh0 c+iQdmuAoBu/pS3tOvuH8TbYbA2ru5bSJ8wtrObQAGaPi4cj8fP/TM/fPjCmyVWRictPA8SmuTtA WIgSKRoliJlcjJ9NFCulOE8YK4VySmm8LUbUuPTHHzKqvWyZfN69F1dp8YFhtO/CKuMIWwDnk6JZ aimidAilSfiDQ9boSHIoimXq1Kki+h0ihg8nGjJEXHyZKDNdXnuRChXRZWuKV4JSbzhL6tFyS1di 9MkToiRJZCTAWVOXGAmLUSZUIKRes6acuqLSuLGMkMLCyQb+hq80Z09fWnXYVBAeM1pcGlRrBeVN V86hS71586YQoe3atQvkselOoMYLXsu3bmm9EtSRTqaZu3oHqiPtWnEaVf5JfreYRIcSCEyICyv+ Bj96/u5hgMg0iU2T8LxHr94/JUfsviJHjCKEJsSlFJkmsakKz1jR7GtwGzoUU7ykJjPnzvNLtHj/ MNp/aU2gNUOUlshWl5oVG0KpEmQJ82dxFFOmTKGTJ0+KiYQhAoPQER0NGIj+T5eVVGhqPSeskAkN qKZCRhROHTA60RO6EqPNm8sOUv2syHNhMcqEGPznRJ5YbUJBPRgKkFSPQRu8+/iShq2pR6dumszT v0+UjUbW30jJ4jrWLgg7UgyYwDS1zmYjCt0NfNWY6gvXrAAbYU359/Y+UUdqPpq1cu42oo50986/ RGPT7t27lR1eYnHfR2XhUZWDFAziwLAPnOOA5KXPI2Oa3FpkE44K5pHE0BIxQiRFaCYTAlMVm5aR zbjfwaE+ZF8u/FZRA4vaY3OaNEHDU9Db1VtPz9Oi/UPp7ysbAolSZAZKZW9ATYsNphTxMob0Yzoc +GFXrFiRdu3aFfLa6sGDiQLM81+n+IHi3D2njx+vLc6ckeVHoRh84U60akX0ww/688HVlRjF7xX+ olevar0SzwcTGfCDLFpU65Uwugc2LbBuMe+eQQX8qlVyz2sH2AkPWFmNHr0yhfiKZKlB/WssoaiR HN8ijo75unXrUq5cuRz+2q4GpbiIZAToO82xVkeaLWUBGl53Hc2YPJeWrlpAKzYsopfvH9HwMQOo QvUSogP90n9nKFGq2PTi3cNv6ihDQ4TwEShejKRGgWktshkvRhJFZjo+TYwoMMQazs3BlGYcn9kq m/7vyb+0cN8Q+ufq5kC34zOVydGYmhQdpBygpXP0sl3Dy5eysx41pAAmlyjr0TsIaWPoBY7+EOL2 UFGKIXj4eEeOaL2SwOhOjG7ciLGZWq+EYRjBixfSvXvfPtNt2bPL/6hp09r1EgcurRGjPX0/y45k pCZbFB+m7HAHUkgjUsFx48YNoyE7LIQwQc0TgJ6ePx/d4FqvxMRnP19RR7r7nMlXNmqk6OTn/4X8 vn4J8+sjUhg3RmKzKOa3qXMI0fDhtOnWDkqM4oABwVJ1QqYt4E6waP8QOnIt8KhBeOyWy9mUmhYb RIljp3HQql0IuurhNwpy5pRRR3eIjl67RoRmPF9fovLlPVKUIrag/k5TptR6NSZ0I0ZPnZI9EPpY DcMwdPYsUfXqmBNoug1+oosWSSsXGxjInxbuG0x/HDR5UUaPEpMG1FxKBTNWdehSly9fTj179qRH jx459HX1QJUqcrBVVcd+ZQ7Bso7UHnAwgtS4ZQOQeWQzYazkitDUnfOgEWtiFL0OmJ0AHRNSLj84 JkTp8Rs7A92OMoMKP7agxkUHiKEErmbNmjVUpkwZMQI7RMBvFNHR9wGWWBCk7nJwiG0Ihnio5M5N NGOGR4lSbE8wGVjZZOoG3YhRbGi3bGEx6irY2okJFsyfhL8QRv0B+AmhKwNml3aAaTS/rm8UKOKT Mn5GUR/qyEYNf39/0aB04MAB2rBhA2XLls1hr60X4C6CfXmHDlqvxDqoIx2+tr5oKAIR/KPSm6cf 6dNbIuVnIM67d+hHJQqWN0Y3Me/eHPz9ENluhYK2AF68eEHjx4+n0aNHi7KLIkWKCAEI7iliAS4J qAeGIETDjbor+/nnn6lJkyZkCepZfXx8hLDC76S2cmClzmA/deoUzZ8/n96+fUs1a9YUp+DAe6Ee Nlo0kwUZjt3Qy4eSitCCSVc4gDt1869At6NZrNJPralRkf6i2cpVDBw4UNkEfKCJEyeG/Ml9+niW YThGQ8KuKqTCXIcg0YXgtZ5S9brxXIAQ5ZHHrgMHevv3a70KRnegbgpeQvXqmYQoaqiUHb+9QvTe iyvUYW6+QEI0f4YKNLPNcYd3DJ87d458fX3pzJkzHilEAdyy7t+3/Tit+DFNCVrR/Rat6HaTtvR+ TY2/n0lpntemO3/GIvwEMM/g5NZ7lDN1MUoaJ+03QhSgC3+A8vv6+tVUS4qJTWqke/LkydS6dWth sg8ePnxIG1EqQvI3gEai6tWrixNspqwBYVujRg3hibpz507xOIhPgK5xCN1q1aoJIYsGuOBAdNdc iAKMW0wXxjLPbCkK0vgme2hKiwPK91rMeDuGDmBCWaOp6Wj6jm6iAcwVtFAOSDEi9MuXkJdezI7Z iz7H1Umhc1hAeQH8UjES0gOEKChQQB48hSaK7yx0lQfRW3cXw3gVsGOBu7r53PFMmaQQxbkdHLm2 hX5d35jefzJNoGlYuB+1LvWrQxtJYNuUNm1aSpYsmdhZejIQo+Ylu3okSsTolCSOnGnevHlzcYKA OXTokPj7QCxC+MXCgY0VsmTJIv6W+5Uj5FLwsiKZIu4VMPMc/qWY3oSOfGtROohZpJNtUaJECYoQ IYKw/GratCmNGTNGjDSF2FWZNm0aXbx4kfKgbiwYIGixVoxFBY4Qoyo5UhWlSc3205nbe0Wj0/m7 0mj2s98nWndsKv15ei5Vy9OBGhTuS3GiJ3LMm1ohnfKBsmbNSlu2bLEZLbZk/6VEFG/cNaqT46r7 pDxRUAlbH5RfIBuE3xTqRn/+WeuVORS48KEWHZFRuOroAV2IUdUpRjnwZRhGCzD2DPWhKOBXQVoK 7t12+UYaaOnfo2jBvkHGdCkaWvpWX0jFs9Z12DLfv39PHTt2FClWiFGIEE8HfoDmtq7uAgQkxB9O sHRC1HFfMKq6YcOGtGrVKiHwkKK/dOmSUZiiHAORUwgja1ZdSL9fC/jt4jdhT40jfkPmIhTg9wUT //w26gNh7QTBHS9ePGEOj3VCjGK6jSPJlaYk5WpRkk7d3K383xpMl+4fFbd/+uJLq49MpM0nZ1H1 fJ2oQaE+dnuhhpSFCxfS3bt3Q/Vcv+jKwUfevA5ekROB8IS5LyKhgwbZfRDujsA4AEM19CJGdZGm V311I3+bvfF6sBHo1KmTsjNy7N4I29rvv3foSzLuyubN8sjfXIiioAi32yFEP37xoaFr6tD8vQON QjRJnDQ0vdVhhwpR1AZivjgiW2rtoDeg9zS9vYSz0U2NyCfqfhFRxTkicRHMZpvHjx+fevfuLepH LYGA7NOnjzjt2bNHpPOXKQdSOKEe1BpIs/ta5CnxfKwjM7o7ggHiOEaMGDRq1Chq06aNiOA6MjJq Se60ZWhGqyM0ptE2ypzcJO4+fvlAK/8ZR/Unf0/z9vYXXr6O5ntlR1GsWDHbD3R3lN8QQXRfvChr Qz1YiALYOh48qPUqTOgiMrpkidYr0CeIEiBlhMJ+NGkkd+DIhI4dHfZSjB6ACES9J87tvYwTiocx vk9No6FLHt3y6Jq3g0evb9KAFdXo1tMLxtt++r4kDamzmmJFi+/Qj4i06uDBg6mRl8279hQxaosU ygdF3S/EJFL0w4YN++YxXbp0ESl91HSa996iaQk1piqPHz+mJxjpF/C61jh//nwg0Ylt7ZUrV2i7 eZmKDRDtxXAFiF68PcQoRtBGdlJkJV/6CuKEcpiF+4fQ9UdnxO2+n31o2d+jacPx6VT75+5Ut8Av 9F0UbesbUXaXLJmmSwgZGTIQLVig9SpcRqFCRHXrysqEiDpQgjpYggwVp3K9a4XuQUoKXaYouLcJ upEQPQ2pGAmpgPHE56NpAlEbd1k/th7OqMFCqBz1ofARtQOkDtFF/dbXFI2p/XM36lB2gsP8HxG5 UhtFVqxYQWlgF+NlxIwpy9dQ0uuASZuagKiorcgoQKr+999/p9u3bwuBaQm8Y1HjiUhkcNHLJEmS 0C/BTAZDw9ukSZNELSRYvHixsAfDxKGQCkmsqWnTlsL1AD/PAQMGiWELiLA6iwIZq4jToSsbxESn /56cE7d/+PSOlhwYQeuPTaU6iiCFMI0e2XqdbkhAkxjcCEIi1PPlC/PbuhZ3sZ5yENiWoGkciQM9 uFbpwtoJ2ygcRYXGPcIbyK6IAxx55wiuIKlyZaKtW4O+n2GCA7V5sHPCZCU7WH1kgpgxr07SiRwx KvWsMpvK5mjqsCWh6WXcuHHit+/tKMeltGYNkYcaBhhBeh2NTEjHjwgYKQlQA/r69WtxGbuskiVL ijGVf/75p0jp4zeyFpN+giGmouqRXQqvKPuECROKyGvx4sXFfWiswv3Ro0cX1/H6s2fPDvK1LK2d UHJdowbRlSuyuQ71qFjf3Llzbab8w46BDl5eJ0TpracXA90TM1pcqlugJ9XK35WiRY4ZpndB1HrW rFnCYovxDLp3l8b3evAb1UVkFMDYmQkD7jDdgnEe+PsjfIbzkFzGzhTmiCNHylnzNpCTd9rQ7nMm gQjfyBH11lOmZI5rVEB0DCl5WwLDW1CbmDxdjKIh6N27d6L5yRw0NKkgwopGKDWOArumqnZMBHin jqe0gmrxZC+W1k6ILqkJLDTWoYwAvqUtW7akw4cPh+i1Q044KpqlNhXJUpP2X1ytiNJhdPf5FXHP O99XopZ77dFJVK9gb6qRrzNFjRTCWfMB4LMsWLDAbjGK3jD0LiEdzOgT+BfrxeJRczGqxmUDDlCZ 0IIvEHF3O8XIX3vDUfKU4SlzlnAhFzDWLvPzXfd8FPhocPBhbSZ59lSFaFjddRT3O8f5CW7evFlE lI4dOyasZRjvqRsFlkIURLByoGSe9rd2vys5fTrwuFZEX9HY1ABWaQHA5gqlBxGdVKAH67QS2epT 8Wx16a/zy2nxgeF0/4WcEPXmwwuas6efyGg0KNSXquftSJEjRrPxioGB52jhwoWFCb4aQQ4O5b+v GKzCYlS/oExW2dTqAs3FqHrQyIG9MBLCOPuSZtLSIXMzJ62H8SjO3T1IQ1bXptfvnxlvq5qnHXWt MI0ihP9WPIQGdZ480psQos5qAnFH3F2MwtoJHqE49wRg7VSrVi1RSgIweh0GFJag414F3qVoRFWt oJwFRGnp7I2p5A8NaPe5P0QN6cNXN8V9+P87c1cvWnV4vPD/xf/hSBGi2vW6iFrDbovxHCBGzSba aorLrZ1OnJB+ok/l5LhA46hQd4P/0CHMmHgsY8eOFYXw6KavV6+esHhyFGztxNjLppO/U88lpY1C FPOyUR/ao9IshwlRjHNUp+1gR81CNDDu6jXqqcDa6b06d52+jYxaA0JUtYJq7QJTbTQRlsvZnP7o cpV6V50n7NZUXvo8puk7ulPDKelo44np9OXrJ6evh9EfiFzD5z2gHFtTXC5GV60imj+fqHdveR0z 0gEahDGBcPRooj17XL0qfQLPO3R9oqP4sqLUZ8CGx0HA2gk+YwwTFH7+n2n8ltY0eWsn8vsqxwHG i5GYJjXbR5V/auuQ93j16pWwx8GUnqAseBj3j4x6MvBIR4VUfDuczPBbx7Y8J4r1AgjNqM2QED5c RKqYqxUt7XKNfqk8ixLFTmm87/m7hzRlWxdqNDU9bTk1S/yftwUcBzBkwBZoSuZeJ/2jl+ioy7vp b93C6Dd4sUkh2rSpjIhOmiR/vCgRg/dsVPsyBwzDOAHMvh60qqZx4guA2faIehsoQUzH+N1iDjma PZDuxFhGa7WCjATZ4GbN5DxpdwQWdZicBG9PgOaea2ZDFhApzIC9opvchzQ93E0QHUWPHTzSN24M 3XczYcIE0eXvTCsocyA4/zw1l5YfGk3P3gYOtyeJk5qaFB1I5X9sLkSsNTB0wM/PTzhdMO7JjRto tJMtCChrhhkP7JtfvSKaOlVORE2d2rVrcnlkFKlh2AlAAkN8wgoD9l7w3QZTprAQdQU4ELhzR+tV MHrk8oOj1G5OnkBCtPyPzWhqi4MOE6IA3c3Tp08Xs8ZZiAaPu6fpMcFn9+7dxuuwUUINonoy//uH 5b5bt4rS06d5HPqa1u5DYx0inAD1opjzHVrq1q0rSrLQpa6KdWcSMXxkqp63Ey3reoM6l59M8WIk Md73+PUdGre5DTWZlol2nl1E/oav3zwfXfVLliwRgtTTuPboJO27uJJgl+WpbNggo6GqDa95ZBRl kpiIiuy1q9HEZxQ1ofgC1LpRpDhg6Ax3DnhuM86nWUADUzNuYGLM2P7vApq8tSN99pM1ZBHCR6QO ZcdTrfzdHPL6sOjZunWrmKaEujt0HTP2gTJauBN5mTe3XWAvhh3pb7/JAWKu3K5VqCDLnsJiT4j/ C4jCQpRed3HOFHZtm07OpBWHxtCr908D3Zc8XnpqVmwwlcreMNAgCwhnDB5A2UFQuKu1U/8VlenJ m7vUovgwKpy5htbLcTjoQYMNGY4lMA4UUVIMHho4UE5AxVwVTER19TRUzUzv58xBQbfpOqKh+JK4 qcYEZipjAwUPPMxpxslRsBhlzPE3+ImGhg3HTXXJsaPHF2M9c6Up6ZD3gLUNUpH4TZcrV84hr+lN IK2G4CK7XQXmwwdplYuID0CasUsX170/mkDQwOSIac3Y1sOAH5xWXhSlAM6ygrLk45f3tPHEDDHr /s2H54HuS5UgE+VJV1YcnEKUvnr5it68eUsZM2RSbosgbgsfcI7r4cKFpz+WRKAc2SNQvryB78M5 7rf2PNP18MHcF3B/EPep7x/UfVJUb/B65AAAIABJREFUB23f8/zdfWo+Ixu9//SWMiXLTS1LjBAj WD0JZKKHDyfKmJFo0CDY6RHBrWvxYqkJcEDnajQTo5jAiNSGmpXAGOLBg7VYiX7p3r07lS5dWjQw NW/eXNRd5cmTxyGvzWLUHTHIfwac/G1exuP9cd3GZYwQnLS1A/17+4DxndInyUkj62+kxLHTOGTl 69ato44dO4qxi5jlzYQcNBxiKFGxYlqvRD+gdAEZtdMm61vxHSHK4woePpRRpqdPbT82pKxevVoR DMOFFRQmQrmKj198aN2xqbT68IRAo349jeDE8Ge/j2K7qJI6YRbqVnG6ww7MtQY9O3B/QAS0dm2i x4+lsxESVVevahMU1MxnFB7FGP9Zpow8ouzTR6uV6JfJyHMEMG3aNOFTZ0uMbj091y4xEjeHPz2M YKDVR4J+DM5xPbSX5Tr8w3DZPtEVVjGmXlbro/Ty+dUOdldTIltd6lt9IUWJaNvY2l7gCoETRj0y oYM76r8FaUaUeJnjSmtAeyydQgtqSWFx1qpVK7HdX4N5sC4gaqQY1KhwfzGtae3RybTmyCTy+agD 7x8Hg1HG+EfflsV+w51nl+mXxaUoX/ryNKzuGvEduTMo+VmwgKhgQaL162VUFAHCNm20y06LyOgD 5fDS3DPNlbRrl4waNHhDxYtr8/6wk7FnmoSW4G+TMmVKMVbO1pzjEsN4egATOpD6al3qV2pQqJ9D Xg+/14LY2hFxfagDgB1ewoR84G4JatxKlJDRHTS+t21LFMxYeYeCKCzKBGBJ6CwwDGLSpEnUr5/8 f4nGIVel7sH7T6/p8NUt9O7jK+W7/ioO2iHkHj16SPsP7KNatWuJ28zvw/mjR/4ULfpXiv7dt/fJ 6/5Wn6deF/cHcR+uGwKeb+0+nOP+oO8LWUI4csQooumrXsFeFC9GUid90/9n70zgp5j/P/5Jjij3 VX6KDuQufpT8Hbk6UcmdXJU70SFHKYRCxy8qR0kpISrpoEgHKiJFORKKlKSSSsL85/mZ72eb77T3 zu7Mfvf99Fh9d2d25jOzuzOvz/vMPVxTHn/c+ZtQSeJH/Qg3SYdS33zzjUWpClPKItf888+eqnTp 9YlXzAJLly7V7c3eC0tz1hhQ7J4fD/26EyFitORC+0P9H//awjHa37r/Cs9T/Huvsvur6896SJ1U NXMXOsLzQfsu/bo95f4sX2sRhRAqjXz7rRMTKWyDXglHHeVUB8FyjOvx4otzs++mTZW64orc7Q+G Dx+uLaY0QgkajCP0qzeTzvzBiimG129crW4bcppat2m12nXncqre8VepK07rrPbfvVLQg/adzZuV Ovpop+Qm1Y3wVgfFjr/++quqVauWmj17duK1Sxhz587Vte/CDO75L7/8Uk2aNCmp9Ruf2DopYbJq 5Q6qXLlSao89ir/OvzxP929n3ztk8HdsoZXM3+mKsTAev1/djXIJVhxiQjmOqdK9wldw00+fnni9 QmPgQKdGInk/iFIeuYKyTrkut3nuueeqBg0a6LJo3B9qELQaEJR5yk8xWqooEar0dgUue795o/rH +ltdc2Y31bz27arsLnsFM8QcsOuuTrISdUY7++MQS5vAe9MLsSHZY+TIkbrjRbLtEds3fiap9SIJ TBdlMkJBKM7MmTNV3bp11X333SdueZ/J91qj2YBWhrjkKU2Ta377zSkSnuvqBuXLl49UWqEqxY8/ /hhYnV4SaxHDmzZt2i7cLR9LO33y3Tvq8Aonqk4XDFa77rx7zvbL50lccFAx9bvscpyqWXNBIPte sWKFPn4RoyHmtttu00WXTyiKkCej8ulcBUMJQpLQSYmONIcffrj+rlIBQvAfSWDanlGjnAQiStTk GpKXqAhTKoDIKCZ69LgnycmAB41wu9KlS8d5p78ccMABWkxEY84cp+xVPonREyqfrR+5Zty4ceqe e+7RFUcKDdqcc/wiRkPM77lMCxWENCD5kdqhtDQEutQI2aFCBaeEEFmvOdQboYY42ocfDmbfhENn K5M+WfakY0wRy5cv1611cd3nshSU4A+77LKLbgtbaHDcOsE16IEIwVCrljQYEDLjnXfe0ZbQhg0b qpNPPjno4ZR4EKAHHOAk7Ag0UaDSiFJB9U+grFTt2sHsOxrEkvbo0UOXgkKUcoPPFeRf/Pnnnznb XzYJpvK6IGK0QMEbQBFtQUiXjh076oz5u+++O+ihFAziqt+G6bQUhJscwYIYDVsDgiZNmmh3PZPE tQS0KieMJtuMHj1adaOpuQuys087Leu79h2+V/bhiCjNMRmJUdzIv/2W+w4NQe1XEAodqm8Y3n33 XXVqPgWElQAkickBQU6xhmuuCWb/CxY48ZA8wgZuz3vvvVftvffe+jkJsC+//HJW90lWPQm3buGL s4TJU75x7bWOkD7uOBGluSSpmFHaktGKksDoo48+WmfQ8YV//vnn1cqVK9UjaVb8JQuwQ4cOahRR 6CmQ6X7zhY8++khnTG7YsEFdeOGFxYLVY/HrrwlX0RB8TwHtihUzHKRQMEyePFlNmDBBx6Q57JX0 903wh333JVEl+d95SYVySlwO8QwH4R22fwY61Cncn4Nja6patZZ22/ft+6K+Zx5zzDG+72m//arb wrOGfS+fqjP88x1C4CnjSv1Y+sxQHB6RGoQVvlBISoziiiNTlgLxgwYN0hYRP2Za+9pX1qulOXpM xo4dq6644gq1detWfTGhwPBxTNfigMAUhOxQXz9SnDsKWSBXvdfDTtDFRYYNC3b/yUFS4TT9F52q ssck1aJFNrcfDEz+rr9eqT59lLr/fmXfi0WUZoOk3fRHHnmkOu+881Tv3r21dcTw119/qZ49e+qg aSx5tK7E2mnaba1Zs0Y98MAD+u9Zs2ZpqyodhVavXq3bmhH4bKCmZsuWLVXXrl31dsjUJR7t8ssv V48++qgWZYUEweinn366Ovvss1W1atXUz5K5IAiCIAg55/PPHWs8LWDDBDoJLZXv+025tNPn9ieC MDLgMu/Xr5+qVKmSdiXjeqfzCn2piSd79dVXtUudQSMqycBdt26dzrxDXNLa7H57uvHiiy/q/ruU iKFm4cKFC7VgPeOMM1SrVq1U27Zt1f77769Fb6FAADrn8f3339clc+omMa3db7/ktr1hg1LUSaYf rSBEgwklnZT4nQZVVFsoDvNxMsj3KrlNYRJCXk7Zskol2QfEd7j/UnUv36uY8bv+448/7HNZNumm KqnAOdplF+eRj6xf7/zeDM2bO5bRLEQ5JAWGOfJlaC5AlzuMVED4HqFTJ6RZZ+yll17SOTgYCVMh 0/16SVqM9urVSxdcJ4GB9l8GxOFVV12l/ya7FksorvcRI0ZoMUo86MMPP6zjTYkzfeutt3SwMz8A d7FcRCtF3s8880z9MCBcv/76a1W1alW1ZMkSHw45f+DYEaOIcyzFCP0qVarEfY+t35OCNvf8qCSj XvBCS8/27dvr3+0e9FhUIkTDApdAQvLoUV+ITJumFB2c6T8fFFw7581TavDg4MbgDzupMWPeU3fc cYc68cQT9b0600nnokWL7PMzQLcqvfxypS64QOl/8w0+21atHHc8bvkgRagBDTZlyhQtHDHQPfbY Y7ZAbp7xdun8FIayXEm76RGdxI5iGXXXFNxxx216tkyZMnq2deWVV9pf8jHqhx9+0KZc+tZiYcFN v2zZMh1A/QkZNC4QsVU9fdWwmlLDEEsrBX1zWTctDFSuXFm76onPPemkk7Q48Asp7SREgwlf7dq1 1apVq4qEqBAmCr20E2V32rYNdgzTp4evpFO6NG3aVH311VfauvVt0Qwnk/vsUUcdpQ1Oc+bMydvS Tvy+OnRwLKFUTXj11eCFqAFPMQKUTk3ucMnvv/9eWzYxCK5fvz5uuCTWTLzUhFwCRkHeDyRLU6Kr RYsW2mIK7733nmrTpo02Mrr36TdJW0ZJNkq2uwotwhBPtCsjAQcQqbwfNc9NbuLEidpCaiBLn4M+ zf72cjJ5DB06VJ8YYlU7d+4cObGFABcExHzlosr0mOf5DEoy/DhwGxmY1JgJCj+Mb775JpTLCCHZ K4d+U8JZuGgY6IB0eFE/RCYu3FzSWbbAvvKybbq3MAkyZLJNs6xRo0ba+iJkBmE1u+/ueEAKLVnR vhzq2p4jRgQ7Dvs2pbP5SwqmFJSJ/8NQhDC9lHTyNOC+jvf06adr+TnMnGFraTVzZngEaDQwCh5x xBGR5+gq8moor3XfffdpwRktXBIxiXUVL/f8+fP1e/mXZeeff76+zx100EHaWk4VJUQryxG/aLjG jRurL774Qu2XbDxgCiRlGaUPLg8vWDtLudLK3M+vvfZafdAtitLrOFgEKnGlFMgl3sC9PvEQ3Lyw ynDyMEUjZG+66SbVoEEDffLNGLz7LYlQr40vAlboGjVqqKVLl6q77rrLt+2TN8bFPSyQ/IZrmAui ebhn6JyPMC7je5rLxDIKWt95553FxuaepDHOdJcxEeD3ipXEr23y2i+//KJdQYI/FGqtUSqKkTKw 227BjYG51a67KlWpUnBjyBbGy4nIQdiQPPsZPU9ThCRl7uWbN2/2e4g5ge9YWIUoghE9QD7J7bff Hnn9f//7n7aaYgBEK4EJlwRCMLi2IyIXL16swynQVW5w1U+aNEmHA2A4IDEd2rVrp2vWkqyOoRHv dlb48MMPrVq1alnxsG8oUV+3bzaWfWOOPLeVc+TvGTNmbLddtmPfmPT7or0H1qxZU2ybmzZtsuyT pP82r3v3my5z5syxbIGc8Xayybp16yxbJGS8Hc6jgc/g6KPbWS1a9LfsL2bG2/aDfPgsBCEMNGxo WePHBz2K3LJxo2Xts49lff99sON4+mnLatky2DHkAu6vT9sHa4sPa+nSpSm/f8uWLdYTT/xtzZqV hcGVQLp06WJ179497jo77LCD/ixskV/s9Tp16ljz5s3Tf/Mvz2HVqlVW+fLl7d/M91a1atUi68+y PxRb1Fr2ZEN/Tn379rU6d+5srVixwjrwwAO322/z5s2tK6+80rJFqlW5cmXr448/3m6/mdCtWzer a9euVlKWUZKPooF10m0xdcePPvvss5HEJvd2yIh3WzXd7wFc+e5t7mpPQ3cpSsdzW0ajWWpLInvu uadO9koXW4RqFwyzKQPntE2byvZ5/UL3M77nnnv8GGpGHHbYYdrVIAhCfAoxbpR6nsRpHnJIsOMo SfGi8eD+SpwgYTbmXkvcYbKtRcnO/+ij0ipbRrRChapFZZIsgRMtXBLvF57nN954Q1tXv/vuu8j6 thDVny+WU8DCiucPd/8w+wdIOFqy+06HlEs7JQslmkp6jGPYQYgSg0s8LiW1DOXKldNfLIeBaiYB MgGDG+CMPLzK86OePXu2r8ll8SDTnbCAQw89NCf78wPc9dzI9i+0IMcsUYhu+n79lBo0KOhROPGi YaszmU2IhTfx8ORxcI0mHjGZMoPr1y9Wc+cuVsccc5hObIpl1BKSI9b5c4dRekMqCZds1qyZeuqp p/RzwicwFO6+++46B4IynSbskfdRBYEcnYPtGS8lpBCiNDtCwGIU4zVjTIwVvpkuWROjcuMJnnHj xukg9E6dOsVd77SilMdPP/1Ui8J8EjpBQ/mtH3NopiLWhwQBPqt8gXJwVNAgdlTIHCyjIZg/5owp U5yaokHPVUk25z6coLpeiYWygg/aSpzrDzGF1AZ3W8pIlsG7RbIMHs9u3fZW99wzWF122ff6GkBe CPGHQnoQgxtNkJL4bV7HA8pzAzGitWrV0p5HoCoSuSgkI+F1BUpqWkWx/ugFsvVJmEYLwPjx4/V9 jvUxLJh9uffrB4Xh6y5QmM0kEqJuypcvr7+4H374YRZHFR1KCvmZoCUIJZVCs4xiFXXlagQG93hX CeyChFJQuHERPZRvMnTv3l0LGZJiTOjdySeXV1OnTtDZ1ySoWgVUDScbxBJ+3tfdz6OFS2LdNEIU vGGPvN8IUQPWcdZzb9tvS3fWLKPZhC81dUcrlcSURh9J9fxUqFBBl9Ni5ksHLG88bzYhNmXatGk5 258g5CuFFDOKNRLNM3p00CMpnHjRRGANpXyQgTJzkydP1qIzVm1i7kWUCwKrqLObkH3yKVwyKcso Znd33UGghWdQVfuJmaMWYqETqzgxcT2U1UoHZrbMmogVERJDqTLKoOQKbgSJunCFDWbQBMcL/lBI YpRyTq1bh6Nt8bvvKlXUgVFwQdlBXLbJNsnAC4ZAMokyQvYgXDJfkr2TGuVrr72ma09RBNVAT3oS ZOJBcOzqZPtTCilDHbjBgwdHihUD9VyZoZ5zzjlpb5cYkSOPPNKPIfrG9OnT9bG6ISmGOCTgXxop GLCcE4xtIASADhU8zAw9GlxYySJNtnYowp0iwbmCz4XfYyxycZ7MOvybTEcObjxY2gV/wMPGPNTV 96BEwvGRRZ9iy+ys8NFHSqG18mwemBOOP/74SMWbZCB+kWsOGd7ue5dQ2CQtmcm+5kZG4VMvdKNh OTdxE29IJxcSFngPHWsQrx9//LFeRneHRx55JPJ+svMQAcDNjb6rt956q+4IA7gB6M5D/APL3LCt Qo01pCwD54OON3S84IdN9wwK5WfiBmE2dUiOa6gkKu1EmQpKVLlLi9Ce1ojGvn376nOBux9ocTZ2 7NjIuo8//rguGNykSRN1AQ2TY/Dcc8/pdnaUsshHcnGeWIdOS8QkI1jffPPNLB6REI1CsI4OHarU uec6MbJB88YbTp91YRsDBw6MtIxMFZJoCAMTD1z+QriFnwXwkw4KpDMSmbyUCHBbTMiyotUfYoiy M3RWQpBSHgBrKm5MMmkRmVhs6MTCzRE3P9YVal4SAkCcIq8jUvmSY7FhG7Siwh3dtWtX3UGATC8D VkC2zw02XX766Sd98zXijfajdKCBhx56SM2bNy+yrEuXLqpmzZr6b8aJ8DNlERDdppYnx21abbGM JCJmj0AMh+lqwTL6xx577LH6OceBADf74zwfU9QKAsGOdcnsz50BhyglS45Mx/r162uLdKbQueH9 99+PPOfc16lTJ2vLsNaZcxsNLIJYIHEHnV3kKyN8hPMHO+20kxZIfGam564XaqrGC7rmfLJNBCld jpKZ5FDaic/7uOOOi7zG52DidPjekknq1zJ+Y7NmzYqcAy+5OE9gKjBgieX7TJu4WPD75bfPemwX txHXiVNOOUUvxxI/d+7cSKkQfo/mN8jNyv0+wnPMb4LsXTpS8TrLuUaZrFGqDfC7MO/j92diqKmd iAg37+M9JoyA6w7n3bzvP7YSMsH8WJiZGJv38fpuRe2A8BKRoWqOgXCKbJayMWI0ZA4M3+DSRh/6 F14IeiQOiFH7tiQUwbWSezj3wHTBuJTL7nX5DhqD627Dhg31c65VVMzBcBcEJlySsAs/SClDhdg4 Tgbtvgy0j6pYsaIWpcAFn7qVzHy4IGOl40KPlcVYvrjBIJqIa+QGyzrUsCJ5hsBoBCsP2oYSGI0o JR4FIcwFnhsJF38+FG6ombikcbNi0TXCjvpaBm7m1Og0y9zuWM4Doo9lPMhEN5CRbrbDMqxVBgSX KXvFMor8G9gXx2n25+53zg0TV4jZX7T4HMbTsmXLyE0+XZjtIIypUWZwj5Mb9DGufml+LaP+aTxw 61Anjc8FYcDkyAguBA8WQerZxfpxfv311/r7g4Bwf86GGTNmaDGHGMMDwITC3SwgGuZ77z4uCj4b aK3n/m5kuoxjxgtBKEYssn2egIsgRZNJOmNf8UDcdezYUQtgLLaMYTdXT0cEMsfH66aFqAHxx+/d vM+IVEBsIsx5neX8Po0YZRLJBMe8j7bCRoxi+R4zZkzkfbShJUsYmDBxPOZ9TEBNj+5+/fppi7l5 HxNSfm/AjfmZZ56JHAO/H4pNA32dmWAbETtgwICIh4dJD9c9I2KxOpttci0cOXJk5H0PPPBAZCy/ /bbQ3v5r9rVkgl7OjQprNrBvBL55H20DTes/elJz3TX7Y4zmuzF8+HA9uTDvo081NQaB0BDq6Zr3 UbuQgtrAPQBRYt5Xr169yOSbSRHfP/M+QouqV6+ulzEB4UZm3sc1v3LlynrZwIHf2//f2/78Z9vj 3UFfG831Fy/aqlWrIhMGStGZXtncIKmHafbHvYeaisDrpjwOy3jd/K6YSPC5mfe5JxK0TUYz2XMd oQi8odw7Y10fkoFra6Lrq7ANftf0nqcYPfdO49WKJ0b5XXJ9y4fk4JTEKBd6LoaIUmO942bMjRuV Dli7cN954UbACaTEAzdILD/c0BCjWDeBm5tboPBlN3Gp3LzcFwiELj8EyksgVtMFAYQlJhrxRJ25 SEcjXvH2eIlXiKBYIN69GNHKjYabISLYD7jJc765iUbjzDPP1A8/l/FZMtkw/XCjgUUPMc96iAlu iO7vBN8lRA83Zm7yXjp37qy/t9zwEBncRIGbI3XzuBmzD2AyRR29ZC6WJBTFOlfc0GOR7jK3aI1G ts8TINS5sXMTx00fLw7XjDnWRZPvdrTvNyASjVD0wjmKdZ4Qe96QHgNiflvTh+Jw7DyigRjkEQ0m 2rHCTBCffBZGxLo/C5olIDqNiDXCCfhs8BiZ97mTwE499RBbtLe2lzfSy921gZmk420x72PiYcB6 zOdv9ucOx2Eyi+XdvM9d3oWJMZNq8z73MTB54Dpt3ufuSU6oFjdP8z4mWUaMIlLffvvtyPu43hsx +uSTO6hdd31W9e49RS/n8zJiFA8a4thMGPgeX3TRRXoZ1jYEvNkfHgV+y8DkAeFv3oeXyhhWuGEj 1M37mDxgpICrrnrVFv+b1R573KyFKn3AzfvwenGdMCKW65fZH1428iyM+OWeZ8aJN4X7n3kfv0fj XSCekt+ceR/XFnOPwwP57rvvRt6HwcDcb3CZ43ky7+NeXbtIQePB4d5r3odxyHiimJzgkTPvM0YY YHumCxPLuSfyPcFgYT6rTOA7wMTa3MuY1JvvK98bvIFmGd4nM+lnvHhMzDKu08Y7xbUJz4hZxgTW XMcRZ0xUzTImQGaCSxihuwQVRjKWAxMntIZZxnllOTBJwxvMMr6j5h7iN4m8WghTvlMYMfCE8Xnx fcS4wO+H79GgQYP0dYpt8d3kd813i98vHmoe6Dk8sZxHNB/Xdq5LTCzp4MRknvPp7m1PuCST8Xj3 70SkXLuHHxQXOnPBYQbMBYAZNjEg7lpi3IAQmOYiiuik9SRCji88BXARg8aFyJeJA8XqSKY+B2+s C174cTBLYHbOFyOekCup8GW48cYbIz8K2Lhxo765cbHjC5cOWEdq59gMkExpJwQrF0ks67ieqW3n hfAFrJtcWLx17biJmZsoIR5YV8x2+b5hieeCQ7wzVhR+ePy48q1zSDbPk4HJDyCW+A0mEqOFjqnR 5/0uMamO1WKPibl7cu7myCP3sG/Ue9jnfvuASmLIeUSD67WxWnpBaMSagHO9jhUawmQ+1oSecAwe 0UDQub1shi+/5HpQyRY8Hex7SIftlnO/MPcML9yfeEQDI0qsyhcITB7R2Gmni9SIEVtVo0bNtFB1 f15M1riPGRHrFvBY4zl2I37dv5+LL75YG23M+4xFH+huhNAz73N/ltx73V4Et/cE7xJCP1Pvg5lw gtf7wD4Qoxg+vImSqcI2mXAhdIxxi9+HEaOE4CCizTJC+owYxaLORMYsYyJjxOj333+vvVxmGZ5I I0ZpUIKwNMv+4wpI5lrnDpNzi228O4g6s4xJlQGDHOKcZdkshxjPq8U1G6HJpJf7GOGMTDIQmkwm mJhwrIhxPNcYxfDCMMFkHe69rAdXX321FtVMtDDQ8B1mcoT4Z3LHJM7tCfQjXFJjK3rL/mLFbWRv z/SsiRMnRp7bMzfLPunW77//rp/bs0rLPkFWo0aNLHsmYdkfuH7dnu1Z9szbsg9AP7dnK9z1rAkT Jujn9erVs2xhGtnuTz/9pMdiCyHLvlFa9kxWv27/UCz7BxBZz/3cnslYFStWtFavXh33GKJhCw/L vpGm/L6wYwtza+TIkWm91/7C63PLec0lyX4WTz/9tNW4cWPLvkAXe71s2bKRvzl2+8dk2T+4yGv2 5MWyL8Axt2uLNsv+QVn2BTfysC9uln3Bizse+0Kkv7e5YtGiRVazZs0Srpet82TWMUyePNmyL9px 17cv5JZ98Uo4ZiF5xo2z7M836FFkh5tvtqyuXYMehcP69ZZVrpxlbdwY9EhKHtzHK1SoEPQwQoE9 qbG6d+8edx1z7X7ssccsWyBas2fPtuwJon6tadOm9m+mqzV+/HjLFqaWPfmwtmzZYtnC2zrttNMi 27AFqNZra9assezJkFWpUiVr8+bN9m/uZmvYsGHWpk2b7O97Oeuvv/7S6/OvPamx7AmLZYtN65Zb bolsi/sen1/NmjWtt956K+1j79atmx57UjIedeye0ZN4hMnXzBJQ2LhAsCaZ2B3AOoeSN7GAzE4w ARuLHTFdlssqgxrH5E0rKmJIzT4pxu6uSeZ+jhmdGKJcFmgPO1gN+IyIIWM2mQp8pu4ZYNhgxofL AbeWG/fnjysDl4H72BNZN0eNGqXda+6abISc4PqKZ3VnRu7uZpFtEpV2MmTrPAEWF+PGw0rE9SEe UtrJf0pqNr19C1EjRyoVlhKUkydTQo/vfNAjCRcmPAd3f7qeIyyUEjOaOtG8Wlhn0V8mXJIwk2ie UT4vHng+ybXBvY4HGiszVk+2g/Yy7+Vfrvemprw7jAj8CpeEpBRctC+bV6wwaLcQNXhv1O4TFKsY a7TkHK/YdD8XIVocvhQIUdz1mNSThVhehEMQQjRRaScDbkvcSt4fGmM3MH7cDu6JTqy+vgbEqPd7 hJDL1zp42TpPsKGkF7jMA0pqS1A8v40a0Zo46JE4SEmn6HBdIcaWGON0YyQJIYpXhUOIDonMxHZi dDLx14QocO8mNAQIqeD6zrrhFeGBAAAgAElEQVSEShq4z/I6nx0xzYQboBFIQueeYe4FuOQJ5yGG ltCTaNoO/AyXzI/S/ELKIOxSEaIEYN8eYANovvDxEr/cRJvxJZowJRJYsSY0iSY6xDQRB50ruPgT E5UM2ThP6UCskzS/8BcKdFDIIUrZ57yFQv50XApDH3pgHkruXoxw2IKHfAU8ou6axsnCdYx8kBtu uCELIyuZeL1axBIbrxZJfOTbkCCMIDSx2IhNDCokaZka8CSEIjKJ+6XyBTkiJkmUewGlDUl65X5M ycgXiuqrmbKSBvOcezfGHNZ1lyNM+fjSfqeQN+DW5QsTLZMf0z7ClS8gpnohNXBr/JhDfynZpyQN kS2aL3CBIuCdzGrBH7gnYB3lq1dSugLRO4Fc16KKUYEza5ZzbnPYYC2vwPtGJQEm46lmUWOxS9b4 IDjE82qRJE6iGV4rlpnQSM4z9wzuUya5DqsqSa2IW6yhZNK7hS7Wah68x11e0l3bHPwOlxTLaAGA uZ2sRcprERdJfIeBkiaIBeJP3IXbcw3ZkYXaSUsQ0qGkuer79QuPVRTERZ8YavJiEXOX84oH60Ur JyckJplwSWI6vfW6jfXSTTJhjm4harbjDa30M1xSLKMFAKU4qBlGoWJqrLm/rJSCSKWvcLZIprST IAjbKElJTPRwwMjiarAXOIjRJHIFCxoEkik9RVkkkpBjxRdi8KAuqqnDKghukhKjzH6o22UgSNZd Z8pPqJWFWCKTWfAXgpdNPbu5cx2X2CGHBC9EgYw94liIfSEOCRDPbmsp8S/pLrv77kcUeXF+bpNl 1FfLpANYqhAj5HaVUMmCtpgGkq5MrchUl+2/fwW1ZMmXuti0X9tkGTPmfE0ECzMlyTKKVZS+EWmW RvadRYvIGMerFPRI8gesZghTYhbptMV10dTjpIYrOQwUbL/qqqsCHqkQRpISo3RxocsMxXjhIB+D aLztqiiIn6jDjJA5Tz1FNygK3AY9EgcmN3T7+MuVkcEM2z0pcRcAT3WZrZMUXStPOaWiqlatk72/ 1fakZ2tG24Rcl3Yie5IOLAYu+lVcQYPujjqpLluyZD/VtOnbqlu3f4sdaybbBM4VBawFf8EySqvK fIdk31deYcIX9Ei2MX68JC6lCtdwSgWR1Eko2KJFB6hrrz1C0cQIgUo5SKl8I8SiFEXv6SBBfc9Y kLmFBchtXqe7C8KB1nNAJwqsWmTW0ZGJGRGdAOrUqRPpokRwLbMj4gMbNWqkOwiRTUfgrWmTBnSn oAtTvLZU1DlDyFITk6w+SguQIEEgNR1j2GeiAGnqnNHJgH/DDJZpuibQJs4vEKFhEqPZhvJrdEwk YxcIfaEbXrduBGwHObLwgCCg9XmZMi+pceP2V+edlzuLr5A6r77qTLDsy2xe06uXY4kcOjTokWyD bs80csqwdGJBg1wg5la88omhaxhCPVb3sJIMyVRUXEk6gennn3/Wwg93JYKT2EN3Iszzzz+vs7IQ nAjTyZMna0FI4Xvj4qePNK2jEJUITNzx1MSkXRftqnhQ8N5sl7ZUtCmjLRWtGU1bOWpgYerHQosr AKHMwbAebgJcASXFuvrTTz9p9yd1wYT0qVixuOhElBL7XWTszztw1ZsWfbjAmZwxeaNkCgWKSRTg QXtYHvx++G3xoIICmZK85mbZMuffP/+8XF1//a7aTSmEl5LgpqcqEOWc2rULeiTbWLLEsdLG6H4q CIFAmBjNbHhku/3yFVdcoVuc5pKkbeYjR46MiEqq+8cDcYlFE+hZTAwa/dPpb0p5Aoram/63ZHDz nD69wDrAjZQODyS2UC+RcgRkd5kbKGLUxJ4Qm8Z26bP6zjvvaIEbq8dyvkFh24cfflhbkmOBexsB YmIJ/y0y/xnBYv52PygqvWbNv2rs2O3fk+iRaNux3pPK+qkci1kv0X7KlCFe9hq9rFQpekUPs2dl i9I6lmwdv/dYTL/obMHvs02bNvpv97Xnxx9PVTgrhg//U+25Z/Te6UKwlIQEprFjHY9FmBrx0FDM vherLJTcFYS0wSg1YcIEnTTmd7gDXm2q7Zii+WgrdFwuSfqIiOt0u+lJ8491k3SfKDK1sdrQBYAO AdG6K0UjUVsq9z5oScU+SCgZPHiwFm5k7XlbIeYbJhwhUVcDwiLcNciSBddYISWwW9ZCteuu9dXf f5dTRx31rJo48Tp19NG/qP/8Z2bQQwuM8ePHa8sq8VzeifCCBdR+/Ft9/30P3dZXCBeE7q9c6Vj5 YzSzCz3/+59SbdsGPYriDBum1OuvBz2K/AfjndRo9Re0gLvEUzrhkrRmHjBggF6HltGUdqTGOJ5q UzcUfUf+Dnz33Xc6RBCNgaEPqyn07t1be6ZNS2k6PpI/kWq4pCHtSxgJHcR+ArGbicQQRVk5eHOw WEvB267KUL6oHxwueUjUlgqwpmLled2+kiTqlx12EAiY5ZPpohTWPvJho1QpS1WuPF79978PqooV 31W1a99jfx+vtb/HFyV+c4gw9d7MDJmJGmEp/JaYmHFhYOLGgwsMD2KtTbIVHgZ30hXeCEp8IUZ5 mQ5zXLfsa5B94SmrvSK57DIlJAfzcS6HCNJ8ZP58pZYuVapZs6BHso3333f60BNPLmTGySc71nvB PwiV5GEaraQaLolQrF+/vu6+hHGRXBT63JMYi7fahEuStI5XGqMgXZ0QsxS979evnxoyZIjeF810 ELPk+KD/EKOQbrhkUpZRbnzeYqd0X8BsjPsd8+6hhx4aaQ/lFke8z9w8Udco64Ptbyg3SlS9u10V N0SzvrstFesjWBO1paIFFvFxKHJOXD5DRiLHbMQoQp5EM7oleds8MoMhfMGcE/NZuYvUuj8bHtzA ypXbwRYf278n0SPRtmO9J5X1UzkWs17y+zk1st4vv6y3f6yX2WKtieradbV9gw/2+L3HguDMxmSD MIC+fftGMvO5aNWpU0l9+OH59sSPVn9OTK39f33BYgaMiOVCI4QH46rPRwsUVtFbbgmXOxwbhlQe EsJK586d9f2gcuXKqk+fPnHXjRYuSRxojRo1Iu1CTYWk//znP1rHmXBJ02YUzxn6wiSYY3XFEkoX QECPkDyO6DQGi3TDJZMSo1gZvdX/KddCmykUOFYXxJJZx1g/4cEHH4zcTFHczZs310kUpiOAt10V it2sn2xbKvaHhQhzNAodS1AYCrlnAtUEqLlqIE6XCUC0WBE6YKRKoWXTxwIroD25tCc9ZVTHjnup UaOUKls26FFlH35j7YqyRowg/eGHZ+xZ7486bojQbCxXxPLxWydkpG3Y/KlC3iYx0cJ6zBgnWSgs UFWOihKEpwiZY891dWvXIn0j+ABeX7cWSzVcEn1UtWrVpPdnwisNeNTc3bbMWNi+KcuYbrhkUm76 aG2oDAhR7zruk+C28pj1vK2p3O2qvOtDorZU7v1x4vJdiAKzGnrJmwfHiLlcXPL+Q0OqiRM559TK VGrVqqBHlBuMIOX7ZZhonwgmPliH3JEuxHq7J0dSxD4c5GsSE8VB6Lbkus8FDrVFcc+La9kf8B7n OCG74Eg1XPLoo4/WpTTN9duES2LZjBYuSStxDBFGaDoetDpx95FuuGSehr0XHlighezBfIZQmKZN nVinQqnRHkuQli49Uo0Ysa0uKxCPCszE8XC4GxQIwZCPYpT74IAB4UtcEhe9EGaiGQXxlpJXQ7gk VXcShUsSbkVzAkQplYzeeustvbxevXo6bMt0EzShmSQfYQQ79thjtQhFvJpQLXf4pvtvQgCoIU+i UyrhktIOIU/wu5RDrVpKVa7s6yZLBMRgU2oGCynFxPm3pGMEKaEvphPavHkv2eLzHDV16gHbFf7m okPcMnFI48aNi4hUIffgpi8ybuQN/K7oEmnf30LDb78RrqLUiy8GPRJBiA4WRz/CJYcNG6Zza7iO m9hQwgKpaW5c/ohOsx1auWJ4wM1f1hXDhsXUrEMcKs8h3XBJsYwWKPSBLgShlQ50wiR2lJht/i0E uFBRSLmuqwtAuXJjVJcuX0ddf5R9YqiQYQLhhWDIR8toGMs58TunKQYhO4I/UNrJ1jiCT/gZLomo NELUgFGBxPJo+8KNX9aTTOFdx/08nXBJEaOCEAW81kz08Ej07Bn0aHKDV5AedNAM9cknB9vi4bnt 1uXCQ3WLatWq5XqYgot8E6MkC5JwRZvIMEFt0ZYtgx5FyUJKOwmpoKUzZYFMRm0hQVY/tbQKkblz qf2q1CGHBD2S8HL00UrNnq1UgwZkmjttC8NUhiYbGEEKuOz32ecL9fzz6+2Z8SB1I/WePOtSS85A 8eR4dYAF/8k3MUqGNVbRMBXpJ/+DmroJeosIgpBFdiSQlTqe7ptKoUB8RaHWTXzqKSntlAzUb6QQ NoW5mzRRBVH6yQhSYkh/+mmaLXbOURMm3K+XeQWpG7qD0MKXmnVCbsDThmeNJNowZaZHgyoV9tdD FZU+DA1YRa+8suRPNHONlHYSUmFHYg1I1xcEITqm9BMt3Imz5W+syiUZI0i3bu2vunQ5Qm3Zspfu i8zrN9xwQ9T3UO+WAsnTp0/XBZSF3GBqjYZdjA4cqBQdpT2V+gIHMTp2bNCjKHlQ2onrpIhRIRkk m14QksCUfsKBQCwUFTEomF+SIei9U6fb1GuvLbLFzhmqSpVx2vIJ0QTprbfeqoPmaTdHhqeQG4yr /rjjgh5JbKgCRpnaooTb0DBrljPZpLmDIAQFJZQoq3T//fcHPZRAICRMxGiBIqWd0qPQSj8hSB95 5Ch19dX728/G6dfiCVLc+N6mFkJ2yYe4UTobIZbDNoEbPJi6iEGPQih0zj777GJdJQuRUlahnwFB SAPy/S691ClTg+uxJEPpuYMPtlSLFkPV4sWvR16nVW8sl71hwYIFuriykD26dVOKq3j37kGPJDbE DjLORo2CHsk2iGGl3un334cvdKAkQJIsMfeSUS8kQ4hyGgUhfzCln+66S6lHHw16NNmFzOerrqIF 7zW6G4cBC+nT9HWMQ4sWLdTMmTOzPMLCJuyW0Q8/pMe1Ug0bBj2S4oQ1hrWkIKWdhFQQMVqgMGul XJGQPpR+IkifDPubbqI6Q9Ajyh60SRwxgk5Nd+j4JkMiQUo/+2bNmqmpU6fmYpgFSdjFKN4DugK6 am4HDjGsiNHbbw96JIIggIjRAoXSTmFLJshHKlRwkiCWLlXqwguV2rgx6BFlh2OOcbK1Z87cQd15 553bCdJnnnkm6vvoZ0zLUPoUC9nBZNOHkRUrlKJYy3XXBT2S4jCBJGkpbDGsJQlKO1EWTxCSQcSo IGQI2bgTJjhlTGh/t3Jl0CPKDrRJpXc3SU1eQTp+/Pi4gvQtyg8IWSHMltEBAxyrelG3wtCAUGrX LuhRlGzwGi1bFvQohHxBxKgg+ACln8jMpZ89lQoWLw56RP5DYXAqCGzZkrogrVmzZuTvMWPGZH2s hQTFC3A7h80qz/eErwMu+jCBJ+OPP5SqXz/okQiCYBAxWqBIaafscO+9Sj38MHXjlJoxI+jR+Ash CWRFv/GG89wI0tNd9a0QpM8++2zc7QwYMEA9+eST2RxqwRFG6+jIkc735bDDgh5Jcfr1c1qShimG VRAKndLdulFwQyg0uElIX/rsQCWj//5XqYsvVqpiRaWOPTboEfkL8XamnBUdmU455RS1YsUK9UNR RtxXX32lNm7cqE488cSo7z/vvPNU69at1Z9//qk7NgmZg7GZGMgqVYIeyTauvVapLl2Uqlo16JFs Y/lypegA/cILSu28c9CjKdnQhI1Y8z32CHokQj4gllFByAJ165bM0k/Nmik1fbrTC92AhbR9+/bF LKRvvPFGTAtp+fLl1axZs9TQoUPVFny5QsaELYkJrwAf7bnnBj2S4mCQp8g9cd5CdpHSTkIqiBgt UKS0U/YpiaWfuIk3buwck5tUBWmFChV0/dGtW7dmc7gFQ9jc9GEs57R5s1LPPRe+GFZBEESMFixS 2ik3lMTSTyar3kuqgnS//fazxa1jourTp4/6l1ZPQlqESYziCqdD2dVXBz2S4gwb5lS7kFj53CCl nYRUEDEqCFmmpJV+Ou88pb791nl4SVWQGr61N3a1rV7+KQnm4wAIk5seVzjxomXLBj2S4kg5p9wi pZ2EVBAxKgg5oCSVfipdWilq2A8fHn15OoK0d+/eavXq1eqSSy7xe7gFQVgso7jC+Z7femvQIynO lClK7bSTUq5utoIghAgRowWKlHYKBnfpJxKB8hUKmccSo2AE6WmYgotAkD5H0F4Udt55Z10WClf9 L7/84vdwSzxhsYzynaBAQtiuLVhrpfWnIISXUpZN0IMQhEJj2jSlLr3UcR3ma6dMWikOHapU7dqx 10FcPv744zpZyXDhhReqVq1aRV0fN/26devUvvvuq0s/lSlTxudRl0wIt+VUEZOMBTAoKOVD8tJZ ZwU3Bi8ffeRUgViyRKlddgl6NIUDSbIHHSQZ9UJyiGVUEAKA0k8I0rvvVuqRR4IeTXq0bOkkhcQD C2mHDh2KWUjpVR/LQlq6dGktRGH69OlamAqJsU+zTpYL0jpK0hKESYgCtU7vu0+EaK6R0k5CKuwY 9ACEYGDWSkKNFL4PDko/zZ6tVMOGTpktKhwQj5kv0B6UuvZ0tIlnjTOCFIyFFEFKwfzrr78+5vuW LVumOnfurKZMmaIz75OhTZs2xZ6725OGbRlOqbNs5Xa56SCQIcZVf+ihvmwuZfgehM0V/sEHSn35 pVLXXRf0SARBiIeI0QIF9youtZtvDnokhQ3WLPQZiU0XXKDUK6+ELws5FkxkENQTJzplq+JhBCkC jIL3MHbsWP1vLEFKlyYEKV2aPv/8c7XjjokvV/+l9VWeLPv77791BQG/xGiQSUzffeeU8fHWnw0a LKL33x9s6EKh0quX4wGi258gJEJiRgsUOgNhbOrYMeiRCEBFoxtuUGrePKUmTaJLUdAjSg4S5N9+ W6lXX01ufWJIH3vssYgghSZNmsS1kPay72oX2Eq9OkGqJQi6Tw0aNEjd7pM58c47Heto+/a+bC4l 2CdzhZ49c7/vWFBHmdDkr77KL49DSYFiGg895PwrCImQmNECBbEThuxbwYGbJWGU9LPPp9JPVGJ6 6y2l1q9Pbn0spB3tGZC7Jz0W0iFDhsR8T6dOndT++++v/161alXM9cjGx/2fL+yyyy6+CVEIKqOe pKnnnw9fOSesot27ixANit9+cwwegpAMIkYLlFNPVapKlaBHIXi55x6n9BPWhHwo/bTnnk4R/GQt o2AE6al8CYsYM2ZMXEFqkpoefPBB9cUXX0Rd55NPPtGPQiUoN/0LLzhJSxUr5n7fsWCChBjyKQJC SAOqFwQVvyzkHyJGCxQyHdu2DXoUQjRIDELcNW+u1MiRQY8mMYlqjkYDQYrFMxVBCi1atFB169ZV 84hnyHM2b96sHnjgAd+2F4RllCAvSjmFLXGJDHqsojvIHS4QtmyhrrBSu+0W9EiEfEF+qoIQQugU g2U0H0o/UQ3g889Tb/2XjiCtXbu2Gjx4sGrQoIH6/fff0x1yKKCOap8+fXzbXhCWUeKFd93VaXMb FsaPd8QQkzkhGCijhcdEEJJFxGgB89prSkmzm/By1FFO6Scy7G+80UlyCiNkKlPA/8UXU39vsoJ0 06ZNkb/PP/98NWLECJ1h74Zi+iRDFSpYRlescKyVuQKraJg8LBx7165KYXAuVSro0QiCkCwiRguY Dz+k9mHQoxDiYUo/UTqH0k8ki4SRFi3SE6OQjCD94IMP1MqVKyPPzz33XHUMtclsTAxpjRo1VM2a NdMbRAmAScE++5DklZv9ffON090oTB3EXn/dSVhKVGpMyC4ks/HdEIRkETFawJxzjlLvvBP0KIRE lCvn1PJEmJKE7tJkoaFOHdzOJBGl9/5YgvR50rQVwuebYuWgYI899tD/Dh8+XL2aRAbVE088oT77 7LNir7399ttq9OjR+u+7777bPs8TI8uWL1+unqSpedH+b7XvsLfccot+DI8SJOtd55EE8RW77bab 6tatW8Jxp0IuXfX9+1O8PzydjWiJSk3RBx8MeiSFzdq1Tme2444LeiRCPiFitIAhA3b+/HCKG6E4 +VD6KZ1EJjfRBOnrr7+uBemSJUvU+1RVj0Lbtm3144477ohb2mnjxo26rqebxx9/XItC6Nu3r2rV qpX6jTRshct7RaQw/4IFC+zfynwdBsDjRFpPeWAdxG7z5s3147wEQXN+l3aCXInRDRuczzpMTTNe fpkJilINGgQ9ksJmwgSlzj47PJMUIT8QMVrA7Lyzk7W9enXQIxGSxV36iaLeYQIxSvZ/JrGtRpDW wdRaBIIUMcojWp3Rgw46SPexJ7EpXmmnK664Qm/rn6IBrlmzRi1cuFC7/GGnnXZSl112mXqISt1R OOCAA/S6PI4ioDcK1EMl259HNMGabXKVUU8ERf36nPvs7ysZSFiirmiPHkGPRMAqGqbQDSE/EDFa 4OCqP/bYoEchpIIp/UTB+REjgh7NNqpVU6pyZaWmTMlsOwjSu+66q5ggpXMTeF31hsMPP1xdl6AB eTV7gIcccoh699139XPCALBy7lTUK5J93HvvvTo5aunSpdu9/48//lBff/21fqxbty7qPmhfiguf x7fffht3PH6XdoJcWEZJEsJFH6bEJco44TGg/aQQHHw37J+V9uAIQiqIGBWEPITST1hGjaU0LGTq qt+wYYNOVnrxxRd1Br23H308N/zee++dcPtYR1/Gn6uUjjO9wmPCobg+BfmJH/VC9j5WWx5Tp05V n376qRauPH799Ve9Di0+V69erR/uCgDR8Lu0E+TCMooblkSpU07J7n6S5csvnUTM3r2DHolABYOy ZYMehZCP7Jh4FaGkQ3Fi8jY84XRCyMFTPGeOEyP3ww9KDRgQfOtDSjwhkMn6T+emtPvuu6u99tpL izoSiP7+++9iy9euXatd9QceeOB276W0k5WgrhFu+GOPPVa74nH7u9uSGm677TZ15JFHqo8//rjY 9qhxipvfMGPGjEjYwNatW/W/hx12mLqTJvEBkQvLaNjKOVH2jDwwWhwLwUH0C9XWqleXeFEhdcQy KqjGjZ2sevveKuQZ3IDxXCNGzz8fV3Kw46EX9RlnODVs04V4zNatW6uhQ4eqnj17qkaNGqmdCXAu IlYiUzKlncrbJ4z1br75Zp1kVCpKMUoSi3r06KEeTmByPv3007Xw5FGBUgchINtiFCvkggVOiEgY IExl/fpwJVIVKnfdpdTTT4sQFdJDxKigW+aRs4G1oyg0T8gjsEDiOsVFSyecn38OdjyZ1Bz1gjC9 8cYbbXH7mhamFLxftGhRRtu88sordZY8/7pxhwRgQSUOdVfaCyknjnWHBL0ly9ofBO57xsyjQYK0 7nws7dSvn1I33eQkPwYNIrR9e8dFL20/g2X5cqVeeMFJIhOEdChlJfJrCQUD7lWytMmSFfITSlsS bjFpkuPGDwLqjWKxxYqWS9fp+PHjdRLShUlUPMetbhKXDGTZl/bEOXB5NNbTaMvDyF57OU0Skgih TQlytkhQ43ONEiWRcyisThSHhBcFDy2BiWPv1CnokQj5isSMChHClAgjpAd5N4cc4rjKybjnBpFr ypRR6qKLHBcqlqtcQVmnZMWoV4hCNKHpduPngxAFYx31W4wOHuyE9IRBiM6bpxS9CsJYb7fQoDQg Bv6TTgp6JEI+I84NQShhkCDOjZryKkGVfso0q74QyEZpJ8hGRj3hOyQutWvn73bTHcsNNyjVq5f/ gltInf33V+rkk51MekFIFxGjQjG40JMM4HMom5BjsIySkBZU6Sf2T7WjorbxQhSyUdoJshE3SsUN RG4Adfy3A7c8TbNatgx6JIUN2fPTpyv19ddBj0QoCYibXigGiQCPPaYUHRnJ3SBDUshPjjwyuNJP WEmwjtKNpWfP3OwzmdJOhUA2LKMkLvncuTQtfvlFqa5dpfJH0FDJjOvKYYcpNXBg0KMRSgJiGRW2 g+xsEmDoskLcoZC/BFn6CTFKVn2u9KEp7TRt2jR10kknRR6dO3eOrJPusokTp/u+TbpJxerklAl+ W0YXLnSsX8QBBw0xyNdfH1xynuD8nq++mprASj31VNCjEUoKkk0vxISs6CVLlKpSxXGLCfkLLjWK g3/8sdPgIFdlMU84QanHH1fqrLNysz+gMD4F7Q377LOPqlq1atrL/vqLkll/28fxmSpT5l9ftmmW kdVPz3s/4fNlIsmE0g9at1bq0EOVuvdef7aXLrSZbdPGKawuXX6Cgwkmk5MuXUgEDHo0QklBxKiQ FJ99ptTxxwc9CiFTcl36iRaNiIchQ7K/r2yBZZn6rc2aZVbMP1fwW8UqTXH6TFmzRqlq1ZT65hun oUFQMI5jjlHqlVecz0LIPSgFSVISsoW46YWkQLzccovSViIhf6H0E4KUBKNp07K/PzL76aCJlT1f ee8951+Og3MXdvx00z/7rFJNmwYrRAGLKAJbhGgwEIN8xx1Bj0IoyYgYFZKCCxFtuE85xcmSFvIX U/qJPvLZLv1EzGqtWkqNG5fd/WQTt2inwwzdrsLMvvsqtWkTpaMy2w6hHcQEBt2HniQ43MJ0iRNy z/jxTriN1BEVsomIUSEp6DeMgCFwXco+5T9YRinLQumnHj2yu698rjmKJ+DDD7c9p/QZXUS//Ta4 MSWDH9ZRLMHEi9eo4c+Y0mHZMidpiUlTGFqQFhrE6d52mzOZ9HTPFQRfETEqpARWkjvv3Pb8q6+C G4uQGab0E3GQuEFprZgNcPPOnJmfFnXOD+eF+pqUxaL/9qefKlWUgxRa/BCjFLkP2irK5LdjR6WO Oy7YcRQqVN8gBrl27aBHIpR0RIwKKYO1xID7HrHxySfBjUdIH9zoCMXly5W64ILslH4i85ltv/SS /9vONsceq9TvvztVCKlt8hgAACAASURBVGrWdEQo/dnDTqa1RhHclANr0sS/MaUKVnuytTt0CG4M hQRWfyamp5++7TWu7XvuGdyYhMJBxKiQEWPHKlW3rnPTuuaaoEcjpANi8c03HWva//2fUj//7P8+ WrRwSsLkG3vtpVSZMs7fWIdmzw52PMmSqWWUIve33pq7Jgle+D7yfRk50mnEIWSXUaOc8l00BKOp gCDkGvmZCxlBHBeuvO++K96e74MPlHrmGed1IfwgOvi8LrvMSTjyu43nOec4ljZKBOUrJO+540fD TCZidPVqZ5JJfdEg4JpBYXvKOAWdxV8SwQLKpGr+/G2vkZxKy1fKmPFbFYRcI3VGhazw/ffODJtM 5I0blXr0UScuERYtcrJjKaSPmK1UaZvrn4QFd3LIIYfIslwvw51OGS8yx3FN+7W/Vq2chKBrrw3v scdbtnSpUnXqFA83COM44emnnaSxBx9MfZv33+/E9xIbnutjIGwEKzqxonhbwnI+820Z1mTKqbGs enVnGVnxDzyg1JdfOrVjiQmWUllCWJDe9EJWwOVDSRZg1r127bZlNKQZOtQpP0P5GEoNmYspM3Nq Gxq4Mcmy3C67/HJnwkDnSm5aBx7oz/74F3FEfGpYjz3eMh58Z4llNK77MI4T6OFO0XvERyrvo6Yq rnEaXPDeXB8DphH6neNtYRxhOZ/5tozi9MTbssyIUeKIEaDEQZcrpwQhVIhlVBCEqCxerFT9+o5F 269WkHR9eu45x8KYj1x4oXODv/jioEcSnxUrnAoAqcb/YvXl83nnneyMKx7EqQ4erNT77zt9zwVB KBx2CHoAgiCEk2yUfiKu2FjM85F8SWLC3Y2rPdXPDEF4++3ZGVM86PBGKA+JSyJEBaHwEDEqCEJM /C79ROHsV1/N37ay+ZLERMwgnx0W0mT56CMnpKZx4+yNKxokyzFJGTPGiR8XBKHwEDEqCEJc/Cz9 VLGiU8A87C01Y0FLRLKQt24NeiSJSbXWaN++TqxmLkspkbnfqJETyyiF1QWhcBExKghCQvws/ZSv NUcBYX744U5R+LCTSnmnlSuVmjhRqeuuy+6Y3GzZ4mTM830gaU4QhMJFxKggCElDhn3PnkqdeaZS 776b3jZI/qHn9bp1vg4tZ+SLqz4VMTpwoFPVIlfddhCiWESJS37oodzsUxCE8CJiVBCElMCKRVIT /6Zj4dxjDydLn6Lm+Ui+JDEl66YnfnfQoNz1oSep6tJLldp3X8faLgiCIGJUEISUoX81NSkp+ZSO Zeuqq5yi7PlIvojRZC2jtIKkucERR2R/TBQSpKA9gpSJjLT6FAQB5FIgCEJa4GKdO9fJgqZ1ZCpl hLCM0gmGTl35BqKNEAMyz8NMsmI0V+WcEKLNmjnW2tGjnaLsgiAIIGJUEIS0oTvTjBmOwDj//ORL PyFEcNXmayJTPsSNJuOmp8D8hg3O5CCbMFGhKxuu+bff3tbBShAEAUSMCoKQEWSY0/eaGpGnnpp8 bct8zqrPFzHKZxGvxx4llW67zWkfmS3okY5F9JhjnESpnXfO3r4EQchPRIwKgpAxlH56+mknI5uY ymRKP7EeFrOPP87++PwmH+JGd9nFSRajlmc0cOFPnarUNddkbwybNilVr57TVWncOHHNC4IQHRGj giD4xl13KdWrV/Kln/I1kYlaq4jof/4JeiTxiRc3OmCAc/6z1X6T/Z5xhtPkAAv4jjtmZz+CIOQ/ IkYFQfAVCuNT+ol/E7nhEUMjR4Zf1HnB4li5slKffRb0SOITS4ziOn/2WcdFnw2II/7vf5W66CKl +vfPbhiAIAj5j4hRQRB8h9JPCJJEpZ+qVFHqsMOUeuut3I3NL/LBVR8riWnECGf8Vav6v0+y82ls wCSDJgmCIAiJEMeJIAhZoXp1p/RTw4ZK/fCDk7wSzVVrXPWsl08QjrBxY9CjiE8syyiJS48/7u++ iA+lO9cbbzif+yGH+Lt9QRBKLmIZFQQhayRT+okSTxMmJF8WKizss49SFSsGPYr4RBOj06c7XZfO Pde//cyZo9Txxzt1Y6kyIEJUEIRUEDEqCEJWSVT6CVF31llOIXTBX6K56f0sck81hK5dlbrgAscq +sILUkNUEITUETEqCELWSVT6KZ9rjoYZr2WUcAksoy1bZr7txYuVOvlkpebNU2rBAqeWqCAIQjqI GBUEIWfEKv2ECx9Rk2zBfCE5vJbRJ59U6rrrlNptt/S3Sd3Sm25yPkP+JcSCcAxBEIR0ETEqCEJO cZd+MjVGKdDevLmT5S34ByWoKKu0fr2TYDRkiFK33pretii/9dhjSh15pOOK/+orpVq39ne8giAU JiJGBUHIOab0U5cuSj34oPNavhbADzvGVc+55bynmlxEYhlxptWqKTVrlpOg1KePUnvtlZ3xCoJQ eIgYFQQhECj9RBb22LFKtWrl9Htft06phQuDHllu+fnnn7O6feOqTzVx6ZdfnLAKxOsHHyg1apTT 0pO6sIIgCH4iYlQQhMAg1nDmTASZEzd6ySVKDRsW/z1PPPGE+szT+ujtt99Wo0ePVkuWLLEF1+3q R1fWzlhb7X6AmrL55ptv1K233qpuueUW/RjuMcXyfpbzeOCBB9QPZPz4AOO5jLiEKJx00knq119/ 9WU/0cAyap8eXeOVOM94rFmj1HPPKVW/vjNZoATUJ58o9fLLTgtUQRCEbCBiVBCEQCGZhkLphx7q lIBCjP77b+z1N27cqAYNGlTstccff9zezm5apA4dOlS1a9cusmzWrFlqYZG5dcGCBWr+/PmqSZMm +nHiiScW2w7vZ52LLrpIlSlTRp188snqt99+y/gY9913X3X11VdnvJ10QIxyXtu23X4Z1k+Eqq3v 1TnnOB2xpkxR6vrrlVq2zHHHS81QQRCyjYhRQRACh9JP6EtE0O+/F8+093LFFVeo119/Xf1T1NB+ zZo1WmyeW1TFvXHjxurTTz+NWEO9HHDAAXpdHkcdddR2y/fbbz9Vt25d1alTJ1uIHaJmz56tNmzY oHr37q2effZZ1YqYApvvvvtOtW/fXl1zzTVqJL0vFWWTpkf+him2shtvK8G///5bzaUtURFDhgzR 7xvtKa6K5batrRrbtGmjPiQ40ybavvv3768uv/xy/Xoidt6ZsToPkpcor0W3q/LlHevno486bvxb blFq5UrHCko7z3LlEm5aEATBF0SMCoIQGmz9pzO+qV8Zi2rVqmmR+G6RYh0zZoy2cu600076+Y47 7qh69OihOnToEPX9f/zxh/r666/1Yx1BqjH43VbFy5cv1/tDEHbt2lXNmzdPNW/eXL/vzDPPVHXq 1FG33Xab6tevnxaY+++/v7rnnnuUZVl6G4/aSq9s2bL6/SYk4MUXX1TPPfecuuGGG/QxmJhRttmo USN1ySWXqEsvvVT/i3vfu+8JEyZokYsQPZLU9gQgMu+80xGliE9bq2sr6ccfK4XRl9OIpm3aVKld d024OUEQBN+R3vSCIISKyy9PvA7W0ZdffllbN1999VV13333FVuOmOvTp892lkf4/PPPtdUTWrRo oQWeGyyq9evX1/GjHTt2VIcffrhasWKF2mOPPdRTTz2lSpcurYXl//3f/2l3Ptx9993qmWeeUddd d53aZ599tDW1SpUqehuI1pWYHItAPCOUTznlFP0gphUmTZqkKlasGBHIlSpVUjNnzlRnnHFGsX3P mTNHLV68WC1atEg1aNAg4bmiwxXWT0EQhLAiYlQQhLyDZKBjjz1WPfTQQ1rwIQzdlCpVSic6IQ4b 4pN2Ubt2be3mj0WNGjW0lbN8+fJqhx22OY+ISUUMwtq1a3UcqGHPPfdUmzdv1n8jcEeNGqWOOOII LYrd24D169ervVx1kcxyROhff/2lrbFmO6fSP9Wz71q1aum42J49e+okK6ykO2P2FARByFPETS8I Qt6BUEQ03nzzzdqyifj0gkA95phjtOU0FRB+Bx100HYi0s3xxx+v3nvvPS0eYfLkydplDwhlxO5L L72krrzyyu3eW716dTWDIqvKSZgi5tVsk+Ss1q1bq5tuukndeOON2lLqhTADROobb7yhRTGxq4Ig CPmMWEYFQchLEHpYPsmONyAg3SKyV69eOmOezPhoy73EWo7YdQteXOe437HOYiHde++9I4lLCFkS o3DNIzC97yfpCff61KlTVYUKFXRMKssQs2eddZZ+T+XKldWqVau0C5/xuPdNeALJTLvvvruqWrWq fr8gCEI+U8oykfaCIAh5xtatWyOJSway7I1LG7jEucWcd7mXWMvJiCc5yg2WUcZAkpKbf//9V+/X vR3v+7FwlitXbrv9sT1c+WT1x3rvpk2b9HqEBwiCIOQ7IkYFQRAEQRCEwJCYUUEQBEEQBCEwRIwK giAIgiAIgSFiVBAEQRAEQQgMEaOCIAiCIAhCYIgYFQRBEARBEAJDxKggCIIgCIIQGCJGBUEQBEEQ hMAQMSoIgiAIgiAEhohRQRAEQRAEITBEjAqCIAiCIAiBsWPiVRyWLVum2rdvr3spC4IgCIIgCALs vvvuqlevXqpMmTJpvT9pMfrII4+o6tWrqypVqqS1I0EQBEEQBKHk8fTTT6t33nlHNWrUKK33Jy1G d9hhB1W+fHl17bXXprUjQRAEQRAEoeQxevTojN4vMaOCIAiCIAhCYIgYFQRBEARBEAJDxKggCIIg CIIQGCJGBUEQBEEQhMAQMSoIgiAIgiAERokSo+vXr1ebNm0KehiCIAiCIAhCkvgqRm+99VZlWZaf m0yJ7t27qxdeeCGw/QuCIAiCIAip4asYHThwoPr333/93KQgCIIgCIJQgkm66H26fPPNN6p///7q zz//1AXzTznlFP36mDFj1NixY9Wuu+6qdt55Z21VPfzww/Vr48aN0wX2O3TooPbdd1/13nvvqa1b t6pPPvlEff311+rGG29UJ510kt7Oxx9/rCv/H3zwwWrz5s2R/X755Zd6nzVq1Mj2IQqCIAiCIAhp ktWY0XXr1unWUJdccom69NJL9b8//vij+uyzz9TDDz+s7r//fi1Ef/31V/Wf//xHvf7662rAgAHq nnvuUXvssYdq3ry53g7rX3XVVeqggw5S55xzjrrwwgu1BXbNmjWqcePG6oILLlCHHnqoevHFFyP7 fvnll9Wzzz6bzcMTBEEQBEEQMiSrltFJkyapihUralEKlSpVUjNnzlRlypRR1apV033uGzRooIYN G6bKli2rxSSW06+++kpVr15di9W//vpLvxcxygM6duyohSh9UOvWravOP/98/TqWUxOzynsFQRAE QRCEcJNVMYoIRUwuX75cP2/RooU69dRTtTv9rbfeUuedd55atWqVjjU165MRb9bn9Z122skZ6I7b hoqYxW3PunvttVfk9dKlS6tSpUpl85AEQRAEQRAEH8mqGD3++OO1q7x169ZaTBqr5ZNPPqmuuOIK dc0112hr6QEHHKBfP+6443SM6E033aSf//PPP3HFJdbTp556Sq+3ZcsW9eGHH6rDDjtML1u2bJkW wlhgBUEQBEEQhHDiqxjdbbfdVNWqVbWARGDOmTNHnXXWWVqUVq5cWVtBSVCqXbu26t27t7Zsrlix Qm3YsEFNmzZN3X333eqiiy7S7ncsoiQxDR8+XG/PLUrN89NPP10dddRRevv77LOP3odZb9CgQXp/ gwcP9vMQBUEQBEEQBB8pZSVZGPSWW27Rwo9/U8W41Pfbbz/9/Oqrr1b16tXT1lFATCI6sYwC4hRR Wa5cOf2cIfLYYQcn3+rvv/8u5rb/448/9LrmUHgvCU48x3UvCIIgCIIgZAeS1W+++Wb9bzpkvbQT YOU0QhRIOLr33nvVm2++qVavXq3LNB177LGR5bvvvnux93sto24hCka0utcxwlUQBEEQBEEILzkR o14o2UR5pl9++UW716k1KgiCEAR4bZgwE2aUD/z000/qwAMP3G5S7uWHH35QhxxyiC6nV6FChbS9 RL///rv2RnGtdkPNZ2L0ly5dqveTaDyCIAixCMx8yMWf2qIiRAUhHGzcuFE3n/B2UaMm8Pfff69e ffVV1aNHj2LLKKFGcwm46667dBgPjzvuuGO77VP71yzn8fnnn2fvYFLAtBFGtF122WW+bJNzyDZv uOEG1bJlS53ISbiSH1CvecGCBXHXQSDyGVDDmWYjeIqIn6f8nRsaiHzxxRdxt/X888+rxx57bLvX //vf/+q4/IYNG6qFCxemfiCCIAhFiC9bEAQNtX6nT5+uHwbETN++fXXDiddee0316tVLvfLKK5Hl CJVNmzbpvx9//HEdgtOkSRPdiMILXdf23HNPvZwH2wwTVPIgnj1TqO5x7rnn6vJ1CMdWrVrp2slT p071YZTJ0b59e73frl27aoFNCNOUKVP0ONxcfPHFuntdqsyfP1/XhMZiSnk9E+8vCIKQDuJXEYQs 484R9JYqy8ayTCCpEAsmzSSA6heE1NApDdq2baurXiAmzWtuEGHx3MHEhrOOmwkTJmihSv3gNm3a aGFDEiPWROLHqcrBfindhqUROnfurNsFE4v+wQcfaNd106ZN1TPPPKNmzZqlhSWCjM5stBMmyRFx SHvg66+/PmobYYTV3LlzdSMOs/8jjjhCjR49WtWpU0eXqANaHCPQsSTTKY7ayXSYMzAGPD8jR46M vEblD/Ae13PPPafHx7qUp0O8kgDAegh9St+x/1q1ammrtfmssVRj6SS04L777tMi3w0eJ0rf0Xgk nlCkux3r8X4qnNDhjvHz/mjbZUJCyTxK57HexIkT9XmURFFBEDJBLKN5Cjdfbp5hh3iz3377Lehh BAY3alyk5kG2YbaXeV3pqXD55Zfrtrzmu4Vr3lS9AMqyIeio7xsN4gh54PKOxs8//6yXI+iwILKv bt26aXfx2WefrRthrFy5UosxrHrz5s3TMeb777+/DgnA/Y0Q69mzpxZC5ph5nffBAw88oOsLs01A cFHTGLcyY4/VRph9UtXD/N2lSxc1efJkLUKxCM+YMUMvQ4hjbUSAItwRqm7efffdyDnDsmyOF7Hp PS7AyshniNi87rrr9HtYj7bIWKlvv/12LVYRrgbc5ghXjgXR6IX16XDXv39/nSAaC8QlrnbzN6If a2m07b7xxhs6ZIPJBCEZxIkyeeCcCoIgZIJvllEsNVxMTftO4IJsLCyjRo3S9UPDsIwbrNvaky5Y Rrgo/+9//yuWvc8Fm/09+uij6rbbblNHH310RvuJBjcibjQnnHBCsdexWp122mk6jgvoZjVu3Dj9 2QQB1h1EwiOPPBJzHc4hlivgM+nTp09WxrJ48WJt/cEylSuw5BlrXq6WZQLijKQUfjuIty+//FKd ccYZxdbhe813DIHnBYsl1jtq/kb7HBFJRtTRYW3o0KFa9LAvHlgBEYCIUqyOiF5jdWNsH330kY6X JBwAMUqLYMQfvwXW5/gXLVqk2xAvWbIksl/Eo3HBc22I1UbYDQKY7QLrEuP6f//3f1pQI/BMIw9v uAEVQhgLYFnGOonFFQGJhdR7XO3atdPJRgh46jNjdaTGMlZfcw65jiAWjXWWaw5j4PeCaPULxhht u1hxsWBzrt2VUQRBEPzANzG6du1aNWDAAO26MZjOSsAs+uSTTw7FMi64XFCNOE0Xd4yd2ZaJscOF GC8eC9GIC850m/IL9k3iBDdOsl9pKsANMZ4Y5WaMS5PGA0FArCE3Om7OfmfkYiHDvYuwIGHO7U4V omNc9YgjmlB4y6QhVjmPiFJviACWznguW75nWF8NXDfcWdq4hU0MKi5o97awZPI9Ie4R6ycTSkQS kz0EHoKO1+rXr69d/u5ELHe5uGTbCLu/i7vssotOQOJc0NgDMY4rHpHoPV5c+4QBYEElbpMHx2zG 4z0urhNsn3rLTHCNMHbvn/G7J/rm/bzP/XqmxNoughvrLt8JEaOCIPiNr3f+vffeOzJz90KwO48w LENE+kW8GDtvPNaJJ56oxTpxWW+//bauj/rdd99pFyAPLDeUbOGmijUTlxnueNqn4prkZsXNnJtg LFiGFeihhx7S+/TC+Ng2lhcEMzcfXJ64Etkf7kK6VyHY2RbJKty4sVQhEnAB8+BGhdUIFyPihMxd bpje+Dw3Jt7MLUYM3ljDWOeD7SMKENCMGVescUOS0cuECPcv65KwQfkwhL8p28NYOf/Auefc4pLk +IxrNZnYuZIMVndc55yfaFnUgKuZ+E8Tb5kuxDPyXUHckZXP5xvrGsLvqkWLFlrU8d3i94QrvVmz Zno5Fkisq4hkvh/Dhg2Lup14bYQTgUsbwciEj+sdFmAvCFSOh8Ye/AuxMukJYSGxie8g6+BJcO+L SeUxxxyjv8v8/oMCSzNhAghnrNNuMS8IgpApvsWMcnHiol5oxIux88ZjIYpwK9asWVPfELmZcuMH RB/uPcCaSUIHcFPFTYq4nDlzZswbrIEbNQ0FRowYocu7uGGciDXcb1iSiFlDECPEsBxi7cJ6yD7Z F2C1xCUIWE5NNi5CkXF26tRJi09zHN74PIM73iwa3ljDWOeD7XMOsdQgGBEoHDOiE4sYiR58Joyf 42LiwTocG+fbxASuW7dOnXnmmTreD/HQr18/NWTIkMhnFS92rqTDdwGvAu5oPkeDiVEFEoT4nvHZ m8lRoiQW9/sNbAMxxsSRkBPKIPH78Da6ACygCEETmkLcJ25jk7nPd4vvKN8DRLQpG+fdlruNMOua NsLu9bzvYdw857gB9znf/cMPPzzyWzEgHvn+Ur4KKykPwh0Q797tYhXG9U8SFIlLTJrMcv5mcshn MHv2bB076j2P0c5pLPgtsA2OlwfW21jbcv9txmy65hEnKgiC4CtWktgXRevJJ59MdvVQ079/f8sW Pr5tzxY01uTJk61ff/3VqlSpkvXPP/9EXp83b17k7/feey/yno4dO+pxGGyhZC1cuFD//cEHH1i2 UIosW7t2rTVnzhzLFk1W586dt9u2G/uGo/+1b8aWLRAt+yZmnX322fq1pk2bWl27drXGjx9v2cLU skWEtWXLFmvu3LnWaaedFtmGLUCt9u3bW7YQs+ybpD6mzZs36++ALYatTZs2WbaItf766y+9Pv/a N05rw4YNVt++fS37JhzZFs/r1q1r2TdiyxaXUc+ffdOzbEFh2aLRateuXdzzwfY6dOgQea8tni1b iFq2yLQaNmy43bYPPfRQa+XKlfpvW8hatsjSf3Mc9o01sp4tdq0GDRpEzi3nJNpnUSjwHbYnWMVe 8z4HeyIQd3mi9xvWr1+/3fKtW7dG3YZ7n+Y76B4323Lvj/XNb9IN31ez3GzTvU/337yfdfjtNGnS JPK6PUm0bEEd87hs8RwZT7zj4jfO9s2Yf/rpJ6tq1ar6799++63Yusl8Lsnifm+s7XrPX7TxC4JQ 2HD/ffPNN9N+f0GWdvI7mSdRjJ3B2+bUDZYHK0oSBZY8Ej2wApGMhCUmGbD2HXnkkdr6YbaLNZB4 ObYDbDeayx9LEw8yp7E+fvvtt9p9SmjBgw8+qLeDlcW813SvMcXPvceZTLyZN9Yw1vkAdywdsYG4 N7FgEsuXLMQqGisX4IZ3u5yzFZOXL0T7DkezfLqtfIkso/GWm4QfN9Hih73b8H5/GbfZllk3mpUV orURdu/T/bc5H1hl8XCQ0MRvgGSjl156KeZxRevqFO24jNs72jkiHMCNd51Myiq53xtru4naMQuC IGSKb256Svh4S5wUCriox48fr0vEuMvgxIOYUgSRAZFmsn8RfQayjYnfw7WOuy9ZEFHEduIaNxCf hwAjaYoH7nRuMqzrHgvxc7xOyRxcn2TmP/HEEzpuDLci8aaAyxyIIeOGGUtouuPNELLJEOt8xAIX LvUlTbiE6e7jPc8GXLQIbCM0yeAu1O+vkDxMrIhNpsYncafUOCVZ0m9o38n3WRAEoRDwTYwiAoiL ygeIPyQxyC+SibHzxnYRf0Xmu0mmIZaL+DC2Q1KDWRdxi3CkEDcCK1GsmNtqQSITllQTO4dFh2QR YiURmaY0DyKXzw+rj4n7pYA4IpMYTMZKxQBeA4Qq4pLEEeJZKSJOQodZ5raiJBNvFs2yE+t8xNo+ 55FYPUSpSYoBk11NzK37vYyb80AcHyKUc0sMYKLPTRCA6hxuy7rf8D01kz5BEISSTikrli/UA8H4 BP3zbzQQDJTRyYcC52S+kw2caWknNyTRcCrdwopsXfPc/bcBtzbvM6483M2sh+vZvT7uYwQRFkzW 5+9o24u1H8blFnAU1Oa5cVOadbBaul2CCFQjbt1/u+E97sxatsPDCDjv82jbiXUs0c5Hou2R4MIy I8CB0ASOlfd794VllP24Kyx414k1PkEQBEEQlE7AxIDEv+kgwT8+kSjGLpqYQWS5If4tWmayW1iZ /cQSR4ni+iBa7CrreGPTYsXPufGWeIlluYy3nVjHEu18JNpetLJd7rJM3n3hxvc2P/AzJk8QBEEQ hPhIaSdBEARBEAQhMHwTo1jsUkmwCRIy3mN1RhIEQRAEQRByR0G66YPq0y4I0fDG9AqCIAhCISGl nYSUISHI9A/PB/huhimxjs5SdOoyUKOWuFYqF9BzPZ/OrSAIgiBkipR2ygAyt7GykuHuhtqe9JJH cFDr0839998fKQ6PCKE6gXmY2piGRMuDonv37pFSTvSG/+KLLzLeJueQbd5www26JeSzzz4bs593 qtDvO1aP9Vzz5JNP6mYEK1eujLxGCS6+j9RypXg/9V0FQRAEoVAoSDc9vdYp7UQf9kwgc5v6mzxM mahff/1V1w/t0KGD7nE+adIkXUTe9G5HGFFDk0x6RHGVKlUi76WgtptEy8MAhewzjb+ldBJ95A88 8EBds5QMenrZ07Oc+qolhXfeeUcNGTJEzZs3b7tC6ZSeokc5D7GMCoIgCIVEQYpRPzGtQI1gHDt2 rC60bsoFtW3bCaSuigAAIABJREFUVhebb9KkyXYlhICi6xSgj0W05d999522sGFFQ8SZrk+9e/dW J554onrmmWd0khbuYCyN0LlzZy2Q6WxE1xgscRSxZ106vVDAu3379urQQw/VnYn++OMPXTi+Ro0a WiDSVhSLHcLT3TaTLkzVq1fXCWxYM0liGz16tA7ZaN26tV6HVqAIdCzJtGo89dRT1aWXXhrZBmNA gI4cOTLy2umnnx75231cdKLiuFh3y5YtWuSbumasd/bZZ6sBAwboiQLHTHFyoJ5oz5491ddff62t uSeddFKij9Z3sCDPnDkzavkpN6buLMdDxyvTnEAQBEEQSiJS2ilDLr/8ct1X3bShxDXvbglKf3cE Ha0Do0HXJgQSgg0LYaLlFJmncxBij/7z/fr109Y2wBKL8L3qqqu0qCQkAPc3IQMIsYkTJ+r1EJW8 blzFxCnSsACRBghMBBDdpBg7ordx48a6Xz3bpe2pgX2uWrVKF9Lv0qWLbquJCO3Vq5eaMWOGXgch jisaAYpw98YWv/vuu5FzhmXZHC9i03tcfM/mz5+vi+sSInHdddfp95j17rjjDj12hH+LFi0i+8Ai jWWZTk1MFryhFbmAzyuREHXTqlUr/dlgXRcEQRCEkopvltFCLe2EOMMNjwsW8UbcLK0m3Tz66KPq tNNOi2rhwsJnRNvAgQN1T+p4y3Fd48rlGACRhsUQUQaPPPJIZP+MjZaeCxYsUOeff74Wowg6xF// /v21lRLL6aJFi3T/eNMLHhCPV199tf571KhR2vLLNuCTTz5R0Rp30RaV7QLrEuPKWBHUWCIpUM/7 vOEGWDoZC2BZRnzNnTtXx3kyDu9xtWvXTv3www/qxx9/1JbPZcuWaYsvILrZV82aNfU2Tdwp1l2O HTp27KgFNuPNBQh0WpWm+p1j/Fh5+Yz9ClfAIk44RKwmBtkiqP2mCsluTCyxSIeNoM4hkz0SAGkt LAiCkA0Ksuk2FjUEpF8YVz1CCpHo7cbEvrAKIkq9JXxwjfM+Hl4hGm352rVri/XEJgvb7TZ3d1fC kjlhwgQtYLGwYbmeM2eO7t+O0EHQ4f4eMWKEFqxua6F7O2TPuzst0ZEoWiki902S1qUIQc5F1apV tRjH9Y510NvRiEkMYQCANRALJyI21niIU7333nt1uAGuf7cwNtsmJpe/jcXaPTaW+ZUclQisvFiK 041Prl+/vp5UfPXVVym9DwFL0hvCm4mSgbAGPut0eemll2Ja+eMRbb9MfoifZlJh4HvO5xoUiZLd 8BpkmmDm5zn0YzzxYOJ58sknay9LmCpSCIJQspDSTj7ATWL8+PHafe120bshYWr48OH6PGXC8ccf r2M6iYEErG6xzjvuaMaFtRZ3O8ITVzqvAxZIrLmIZIRrLIgJff/993WYAMk1qYRj4MJHMGLhGzZs mI759IJA5eZMPKUhlljkhoi4ZlvE43pbqrIM2FalSpWKtVINApOElUkdURLZUvU6ECbBxIhyUQh8 4nj9gO+LO543ExBSQ4cO1ZZuA/HLCxcujPs+hDZegiBAQBNLnQl+nsNkx0MYDh6IVGHb1157rRbQ YbQWC4JQMvDN35NvpZ2Y7WeaTW/A3cv2vv32W32jMWAVNFZSrJlY87jxmn7r7uXRiLYcVzUxoyQ2 sU36yZvEH+/6WEARgiQMAXGfWOkQAEBiFFYgrG9sxwg3b/93bpxHHXWUFsLckCpXrhxZbvbpfY95 zVhx+/Tpo0taGQGCpdSAC/uVV17RljwTJ8o54hi9x8X+sZpyTMRfkuzj3i9CnRsnopWEqmjH432e TbCMcnxBwHcF6zAJb1jImzdvrl8nhnjw4MH63N13333aaozli+8C54UQBiYHTKAIu8BSSejDnXfe qVasWKEnVHw2xAlTjgpBRBIZ8dOc/2jJZbHgOzl79my9D++kiu8L35vFixfrz5zvLpUr3n77bV19 gOMiLpjJUbxEPSzpyST9eSdKWOiZ9HFcieD8cl1hctSmTRt13HHH6dd5DWsvvy3imPHKELYS7xwy ZsbK58RvDu+IuWbEg3PPJA5rpjtRj3P4yy+/aBFPgh/7ICabz5ZzjNg85ZRT9Fj4zeCFwINCmA7x 8Hg5GD/XrmgJj8BnzsSY53fddZf+bXLc48aNU+XLl9efidujIwiCUAwrSW6++WbLvkDGXG5fyPCV WvYFNfLaqlWrLPtCH3mEZRnjXLBgQbKHnhT//POPZQvyYq95n8O///4bd3mi9xvsm731xx9/JFyf 19z7tG+u2417/fr1xd7P+rzuhc/YLDfbdO/TvhEW2y7rjB8/3mrSpEnk9YceesiyRXnM47LFc2Q8 8Y5r7dq1evvuZbaYsebNm2etW7eu2Pi9x+MeZ7axxYhli4OMtrF69Wpr4cKFKb3HFu+Rc9OyZUur R48e+m/OUe3atS1buFi2CLFscaRft0WPZYsM/ffAgQOtm266yXrzzTet888/37IFqDVx4kS9rG/f vpYt+PTfLLNFoPXxxx9bnTp10q/Zwsf69NNPrblz51oHHHCAHrvZL5+Nm9GjR+ux2ZMHyxZD+jVb 4FiDBg3Sf1922WVWr169rM8//9yqV6+eNWDAAGvlypWWLSStO+64w/roo4/02A488ED9+dqCU/+2 X3jhhchxjxo1Sn9XKlWqpPfHWO2Jo2WL8ci42PekSZOsH3/8MXJ8toiyTjjhhMj43WNu2rRpsdds 0WrZk1A9Hlu86fHYgtOaP3++ft2epFq33XabZYtA/ZtNdA7feOMNPR5boFoNGza0nnvuubjn0IyH 7bLvYcOGWbY4tCpUqKDPC+fPFoR6u0uXLtXn47DDDrNmzpxpTZ061Tr44IOt5cuXW7YItmwRqcfD /rlmcg7tSaL+/rGOPUmxbCGr98VnAsOHD9fnatq0aXqs9uRAnxN7smvZoth6+OGHLXsCncS3VhCE fIVrFfeMdPHNMoqlgkQSYhoNWNvsG0vkubu0UZDLmPn7nWwVzcLpjY0Et0Uu2vJE7zdwbN5SUdHW 977mtbAwbpM8ZNaNZTnkMzbLo23fHZdpzod9k9IWGRKasJbwHcFyGQtT1ijRcZkYVvcyMy7iaN14 jyeXCSBUEsgUQi0Ik7Bv9Cm9j3OOJZNwBWIzDdTZxWLG98dYA7GCETuMxZmENRo3cG6xSpLg5k2g wqJGmAeWRr4/WBchXnJZLIinxnrnDiXAIkmtWSyYWDWxEGIRtUWy9mhwTMYLkShRD+tcskl/gFUQ SyfvTzR2wNKPhZnx8OA4sBLyPcQ9Tq1gzh+hJe5qCrHOIcdA1Qysm8RbuxMLE8FxexP18JBgscV7 QwIUvz8SFtkHcC4Ja+EcMA6s4uZ3xXr81ox1P1rCIxVECLXBY8MDmjVrpq2txDoT5oPlnc80Wnk7 QRAEX+/KXKB4GBA+bre1mzAtE7IHmfPEAOIm5AaXTVcdLtGSGNdGfdJ0Eu4QFSS9eeNqjdDA/Wpi j6+88kot+HCx4iY2LnOEFhUKSICbMmVKZBsmkc5MZAy4xNku2/Iml8WCiQKuasShPbvWr5nYapPc hOAknCQaJlEP4cM4iYl2J+qlkvQHfGdxY7PvZMQo23d/79g+sdWIMWIuCQsgdtob5xrrHBJbzrqE MCxfvjylLPZkEvUQoXzubBsIdTChPEwGY02CmWRwbvkc2LZJMETwIpq9+yDx0eyD40km1EAQhMIk 3HVWhBKDKT6fTXJVqikViCPGWmRuyumApc+0X00FLF6JrO8GPh+spcRlmnhKGh8gUrBQEgOJhdKA hY2ENiyntDeljBcijoQtxAkiiMz0ZMFyifUNKxuxqohABB7nju2DqcOLdc3tgUEgIagQR+5EPaxz gDDGqmwsc/GS/oAJNesjrLG4uitJRINzg+gkDhprJ7GbnEcsuZxLSrrxWXh/A9HOIeeACUC3bt20 iCUGNhlBnwhzztgn54PYUMZoyq0BsazxcCc8EpeKpRcQ/Rwzx48A5cE5QWhjyQaOM1dx2oIg5B8i RgUhi+CiRYjgujUJRKmAUMGNnWrHqFgi1J0M5k14I5EFAWfKDlGuzCS0YPnC5WxCHngfSTYIJuqn YlEzNXCjJZdFS8bzvkYFAISkseSyb7qEcQ4ROLiIW7ZsqerVq6fDH0hmYp+JEvVSSfozx8c+EJIk I5FcZ+C4EJ4k8QFCk8REzhsltBgnY6TOLUmdJEjxGuESJAhNmzYt4Tlkv4g4LKJ8jtES+aKdw3iJ esaiSWUFQgrOOussLUo5Dqy2JBuZpEM3ZpwQK+GRsAfOOdZpJjBUFeE1wiI4J1hESWLC4isIghCN UlaS024yncmo5l+hOLikuGhjZcD9l4xrLxmwVESrPeqGuDJcuEuXLtX9zsNeVLwQId4TIYrlK1cZ xViioglS7+vu58QNEmdIdrsBdzNWThOHy+WChxEovB+XOuLEwO+B9bFUmm0nOx627RVEFF1ne243 LxZItm9ijNmOW0wxZq9bGMsor7vjNqPt3318CMpkf1OcB7ZttkccrhG1gPhDkCEuE51DriMsJ+SB 4+TvROcw0dgRxcR9m21wLnjNfb3yvod9u0UtzxGchBZ4x0MFC6zIbsGMAOe9Jt5cEISSCdVA6IyY qIJKLAqy6L2fcPGnVAo3R6wx1PLzq5g31jDT6jIWuM2wbBBrl6g+oxAMWOxws7uLzycC4YSlKV1i WUa9r7ufY4k0yS8GxJ47IcxtKTPvd4soQJCwnnvbyY4nmisXseQVllhP3clu3kYM0eITcVV727FG 27/7+FKZ3CHQ3NsjEal79+5ajGJV5PeMEE3mHGJ1RIiCWTfROUw0dj5H9zY4R96Js/c9XmtptIRH A2EVXsstVnURooIgJELMaBlC5jEuKNypWAaI/aI+IS5CEkFMhjC1+bBg+tkYgB7tJElgzUAAmNqG QvjAFWvAGuXN+HdDshduz06dOuViaBFIIpJakP6BNRz3OJ8nQi3oBgyCIAhhRSyjGYIbE4GJwHCX h0FMfPrpp1GtoVi9uPFjhSL7F1eWYciQIToOzdsxh+xeOg4Rw2Y6IBG7RokaElxoCZhssooQLGSs 81kyQeH74wZLGhMahEw6MaaZQAJYvCYMQupgfaQUlQhRQRCE2IhlNEOuv/76yN+mzSbg7urRo4fu POIVpMSSkXmM5YsYMjrVkKlK4D9ZvAhVXjfZrcTgEYeBUDWdbRCkiBpAoAr5A4k6ffv21R1sKNs0 b968SN1bLKZ0qSKhRRAEQRAKATGDZBGyoAnyd1s5SUx48803dXkU4scQrDNmzNBJAdTJRLzieqeX O2VYAKFqilTzflOkWshPiLmjfBG1MLGKu2uIUjT+/9s7Eygvp/+P36gsRbIWWUq2CGUpIrIlJXIs oYiDokKIshxjOcmakqUsKfuS7GuiyGGmJkvp2OqoSAqdItIc939ft//9/p555rvNzHf6zjTv1znP zPf7fO9z733uc+/n87mfuzwyRIUQQtQmZIxWIWEzb95yw7xOwKBkEUVYYMF/FmKwAIqh/uiehmHI NLpJNUd0k2pRs4kvOhFCCCFqG9KCVUx0M29gsRMwFAtsqs1KWla18to8vKThdzYPh/A2G1brs/9g v379Sr3pSgghhBCipqI5o1VAus288YQxL5RNshmO5a0o4e06zCHkHdZsfM3+omGTcRZIJdukmoUR QgghhBA1GW16X0Vks5k3w+/JXjXI/FH25ovHkWyTaiGEEEKIfKJN76sp2Wzmneqd12GT6HgcyTap FkIIIYSoycgYFUIIIYQQeUPGqBBCCCGEyBsyRoUQQgghRN6QMSqEEEIIIfJGzrZ2YuU3r7ecN29e rqKscnibEa/zPO200/KdFSGEEEKIWknOjFG2HLrvvvvMc889l6soq5wVK1b413HKGBVCCCGEyA85 3fSeLYl69OiRyyirlMLCQlO/fv18Z0MIIYQQotaiOaNCCCGEECJvyBgVQgghhBB5Q8aoEEIIIYTI GzJGhVjPmD59uvn+++/znQ0hhBAiK2SMrgNYtf/777/XmnTLi7XWLFiwYJ1dt76z//77m65du5o+ ffpUS6O0qp9bTan3QlQX2JLxoosuMosXL15nabLzTv/+/RPH7Nmzqyythx56yDz55JOVjueLL74w Y8aMyUGOcsvcuXPNhRdemO9sVAoZoznk+eef943q0ksv9RV29erV/vy4cePMnXfeWeF4Fy1aZHr2 7Fnu61KlO3ToUPPmm28mvi9cuNCMHj26wvmrLAjAo446Km2Ya665JiG0BgwYYH799desrquNNGjQ wIwdO9ZMmDDB7Lnnnjk3Sqk/PIfBgweb999/v9zX5/K5JVNolW1vQlRnCguNad7cmCuvNGb58tzE +cgjj5h33nnHy4xMtGzZ0ixdurTSaU6aNMk0atTInHzyyf7YfvvtKx1noLi42HTq1Cnx/bDDDjMH H3xwpeNFjlJOmchVGWWbxg477GDOOOOMKk2vqsnp1k61nZdeesnsvvvuvuLTE5syZUpO9l3daqut zLnnnpuDHK7l3nvvNePHj/eKe8sttzQ///yzefnll72Rlwoa95VO+n3wwQc5y0d5uOuuu7wQ2HDD Df13hNhvv/2Wl7wABtlOO+1kLr744pRhvvzyS3PZZZetw1yVBuH+008/+WfNceKJJ5p77rnHC7HK cMcdd5j33nvPex8vuOACb/ideuqpOcp1WdKVNQqtRYsWCcWTS4UmRHVkxQpjGFhwTdlgOxYUGNO3 r1PmFdTmjFS88MIL3iC94oorfMc/8NVXX5kHHnjAv9QGJwvtHkcAbfLoo482Z555pvn333/9HuOf f/652W233cygQYPMZpttZj788EPz559/ernNaA0vmInTunVrc+yxxya+//HHH+b+++83119/vf+O Tt1uu+1Mhw4dvOw65phjfEd7k0028WHQAytXrjR33323NxQZEerSpYu5/fbbzbfffuvzHGA/cfRz uvyuWbPG6zqu7devnznooINSlluq8CNGjChTRt99951P859//jHnnXeeOeSQQ3y+H374YZ/2Z599 5m0G7u3jjz/2Oh99u8suu2R8DptuuqkPw71QPvDGG2942bjxxht7j/e+++6bSG+PPfYwL774ojn0 0EOrlzfVZskll1xiR48enfJ3ZxjYxo0bZxtdtcBVAOsqT87icz0T+/TTT/vP8+fPtw0aNPCfnfFn XSO3w4cPt+eff74tLCz0512ltXPnzk1cz+/OMLQfffSRdcanL3NX4eyKFStsQUFBItxTTz1le/fu bW+44QbrGrtdtGiRHTJkiO3Zs6e97bbbrGtsiXQ5H8dVfusaoD/g008/ta7RJH53ldj26dPHX7ts 2TLrBIQ9/fTTrVP0duDAgT7PzshKpOMMbusEjv/8119/2WuvvdZ/Xr16tXVGpO3Vq5e96aab/H2A M2jta6+9Zp3xa50AtM5gsrvuuqv/zQkUe/XVV5fJ8wYbbGBLSkpKnYtelyzfIa13333X3nrrrf63 OXPm+LySp0cffTRx7d9//22HDRvmy3XMmDH2v//+8/l1Qs6+/vrr/lonKBJxtm3b1nbs2NEOHjy4 bEX4f6ZOnWppYtXpcMa8dUrB329FiT4LZ5j6sgnl4oSbPeecc3yZBZxRbp2g9r85oZr2ecefYaay jra5QLTep6qD2dRfIaors2ZZ26mTsyLN2qNVK2vfeqticTmjyrqOqv/sjDUX9yz/ecmSJV7mjxs3 zk6ZMsU6Y8jOnj3bNmnSxL766qt23rx5Phx6B/1GO6eNHn/88f487dAZVfbxxx+3M2bMKJMubRf5 +s0331hnzHmZQpx77bVXIgw6inTBGU/2iCOO8Pl1Bp3XH0De+/bt69NAljgD0euV/fbbzzrDzh/o U9dpzphfZ/jaCRMmeJnStGlT64y/Unl2Rpzt0aNH2vDxMkJ/OkPR6/XJkyfbZs2a2YULF3o5iI1A 3t9yD49zzhC3znD1cXfu3Dmr58D9UcbBlpk4caI98MADbVFRkXXGvM/j4sWLfXrOcPXlNn36dNuy ZUuvo3LFCSecUErulxcN01cReB2jHiiGDvHa0HM56aSTjDN2jFN8vjcE9HJGjRrlPZX0pJxC9N5Q elL0aJ544gkfjnkv9AKdUWt23nln32OiZ+YaqXEGl3EVPuNQC2lfd911xhm1ZV7fSk+UHhjpb775 5t7jxcsMjjvuOLPNNtv4oQCGBEiTtACvJXkHPKdOuPjP5B+vK6+JpYfsDFp/nnk3DB27BuN7zIFf fvnFl020pxyFMuDA4xwnWb5DWs44Ms748b1whoed4PC92JtvvtnMnDnTh6O3WrduXe8VIC6eC+Xu DH7z9ttv+x4kHsFp06YZJyz98PcBBxyQuKeagDMizdlnn+3vkR5zLqCe0yMHPA2uA+U97NRPZ0z6 eu2EvWnXrp2v16HOQPx5J3uG2ZQ1w/54JvA+4DmIkqoOZlN/haiuIDYRg5MmMVxrzNdfG9Oli3Ft be3n8oBMDdPAaKNhbqXrxHv5jKxm5IGRib333tvLDoa8mzdv7l+p7QwQM3z4cC9feaMhMpK2BsRL G6T9vvLKK17n4LELIItpm0z5QVZkglEYdB3yGNmDfnSGnJfNpIFHFDmOJxAZgkeVwxmA/vpM+e3d u7c/KAfkZabRt2Th42VE/nbccUezfPlynz6jPEH2kEc8wchI8ugMUx+G8GFqVabnwP21adPG1KlT x4d3hqn3GqNfeQkR3ll0GKDD8dDiFWWkrCrn6ZYXDdPnGBoF80VRxI899ljiPEMUVFqg4VFp+d6+ fXtvXOI2R1HyRqiNNtrID22g0Jn/hzINMJwycOBAc+SRR/ojQEVHIWN0ZTM/kGEA8oGbn6GZAIKI IQQUMkbAjTfe6I3XaOMGjAjmndJIyDNGMYKBc/wWGj1DufXq1fONfosttigjpID7W7VqlXE9q8RQ TDKYH8MwfRiWiJIs3wzHQK9evXx6zOGlkdIxoOFiBM2ZM8cLJdfDNGeddZaZP3++cT1qLwC4j9B4 ITTejh07eoMcoUKDTwXx5GtaA+VPncNQQ0hSpsOGDfMCOxdQFjw3yiBMRbj88svNjz/+6Oc4b7vt tn6R0tdOMwZBCgjTVM872TNs3LhxxrJGoaFQ4MEHHyxVBqnqYKb6K2onY8caw1T/khJj6NfEj2Tn sz1XVdevWfO//DOdkZHlqVONSSOaElDn0T0MEzO8zGu9aR8Ydego9Ek60DvoKNoX8B/5TLzAEHQA +c3iwug5hqIx5ALIX/RNKsI0LXQk8p1hfXQZuikbMuW3bmSuA8bemmjhJiGb8KRJXlmbAeijoEdJ O9wTshPZg2FKXKEcsnkOUSgTnFoBpjIgb+P5pQwz3d+6RMZojsEAwKiJVgZIVmlRsK1atfKLQJ59 9lkvDDCSmDOCgbrPPvuYiRMnmiZNmiSuTVYx6dmihLt16+YrPPNisgGjFs/TjBkz/LwhoOEgkELD Id7QcKN0797dHxjTGBM//PCDn0ODEXfLLbeUS0iFMqFniFDEC5sMjObQcOOky3coe75TvqEHyfcS J90RkIARBczTQSBEr4XyNl6EQLTDsC7h2eJpwCuMAR68l7mC3nzTpk1LeViZk0UZYYTj9adOpRKk yZ53tnUvTlyhBdLVwUz1V9ROXN/UYAPQ7BE18YPqHj+XLGyq67MNm+31xcXGDBpkzKxZa/PfubMx OPmzVAG+s4aMYlQigJeN0Se8b8w1R0YiB+mIo5PovGHwMJcz6CZGoGj3RUVFvgO59dZbl0mL+eWZ wLBkxATjCfnBNnXpZBd5YDSE1eTospBH5BB5jFOe/FaGaBmRDnM18eZSjkHXxncuwINKhxuvLSM3 YYQzm+cQBccRzqzDDz/cyzrkWrWaG5oCGaM5hsYUN0TTgXcQbypKG68Vxg7XMxxBbw9PDR7SABWT ykVFQ3Fz4JYvKCjwSn3IkCGJyp4JGizeoquuusp7ooCKzD2ExSI0dIy3eONGQHAeLzBpMykaA5rh BfIf8pBto8d7R4NjAjgGQqqh+lSkync2kCfyjFBGoIXrlyxZkvKaIAiqIwhwvBsI6FwboQE6UtGO Ad7HyZMn+3pMHcb7DMkEKWWd7HmneoYVLet0iocjXf0VtZORI/Odg+xYtsw4uW0Mg2+IWtcczYgR xpxySvniwQkShpgDeO6eeeYZP7KH0UMbRv6jq/gevHcYl3R0Wfh0ikuYziXtlHYN0Y5/Mkgzmi6g 80if9ooswDEQ4oiGD5852AkG3Uf6dDaRQ7RnZE7btm19x5lwIT/Z5jdZ/qN5SBc+XkZMD+OeGLZH r7BgOOQpgAxE73MdcopFWkBnOdNz4HOIiyl43B9OLuwDHBJ0MDB+o+nF08872U4u1QKmzLhGbV3j LnN+5MiRdujQoYnvTBxmoRKwYMI1QOuMQv99wYIF1ill2717d/+fyd2E5RpgYZSrlLZdu3b+P5PN mazdokULPxG7a9euicUb8XQDjRo1SnxmoY4zwvx14HqltkOHDn6iuGsEfuEHuB6WdQardZXafvLJ J/4cCz2IiwnjPP969erZUaNGJeJmkRL5YvFJ69atE9fF8xW9PyZdO4PALl26tFSeGzZs6NNncjsH i6ii16XKdzQt7pXFW4H+/ftbJ3D9ZyaQOwFmu3Tp4ifKOwFVKv5wv8QH06ZNs85oKbXwq7pAnapK eM7xxWTQrVs36wxLXyauY2Vnzpzpz7MojIUR1AEWE6V63qmeYbqyTtbmos88VR2ETPVXiOqIUwmu PaxduFS/PvWYNl+xuKj7mc6zSHbVqlWlfl++fHkZGcBCnSjI2/gCoCjJZEhg5cqViev5nyx89Duf XYe4TPrhXDSebPObqmxCupnCx8uIBZNxvRa/hvicAVnm/iDTc4iHJ574uWh6ycqkMlR2AVMd/mRj tLKHH0M2NisUAAACdUlEQVTK/E8GnhEW7NSkzaYLCwv9Ygv+5wI8OcmGkSlijtCjCl6icA29P4bm WZAUzlGOwXsTvwb4nflvIU7mx/GZnhNzTfgcTzddPgkX7SWxeIfvLF6KhmHok15bIJqveB4DXENe U5VHtvEkIx42nu94WtF7p5zivVvm+jK8HoZ2o/HHwzMEwrlkc1jXZ1LVc+BZU36USzQMw/Y8g9Db T/e8k9W9VGWdqi7H61e8DgYqWu+EyBcMybdvbwxb9TIkX0WDH0KUC7bVYgEr/yuCJG8OSaWg4wZP VOExT4Q5dcEQDfGwcCZKXEnGhxKDkodUwwjp8hkPF5/TGcJEDdF4vlIp8rgRkCxf2cSTjHjYeL7j aUXvPW6kQ3waQTT+ePhcrUivaaSq5xCedTwMczejpHveyepeqrJOVZfj9SuZIZopH0JUR9q0MYZ1 oFlMpxaixiDpm2fiq+KFEEKIdMgQFesbMkbzTHQoUgghhBCitqFN74UQQgghRN6QMSqEEEIIIfJG zobpwz6U7KNVU2CFbthXUgghhBBCrHtyZoyyWpYNr9lktSYRfbuREEIIIYRYt+R0ARPbDentJUII IYQQIls0Z1QIIYQQQuQNGaNCCCGEECJvyBgVQgghhBB5Q8aoEEIIIYTIGzJGhRBCCCFE3pAxKoQQ Qggh8oaMUSGEEEIIkTdkjAohhBBCiLwhY1QIIYQQQuQNGaNCCCGEECJvyBgVQgghhBB5Q8aoEEII IYTIG3WzDdisWTMzYMAAM378+KrMjxBCCCGEqEEUFRWZgoKCCl9fxzqyCVhSUmKKi4srnJAQQggh hFj/aNiwoWnVqlWFr/8/OUT30JtVx5kAAAAASUVORK5CYIJQSwMEFAAGAAgAAAAhAIaewjIDDwAA Jz8AABEAAAB3b3JkL3NldHRpbmdzLnhtbMRbW48bx5V+D7D/YTDPkabuVU14HNQ1sWFtjIzzsm89 ZEvDNW8gORorv36/5sVjSV8bwQZBIEEi+7CqT53znWtVffOnX9arm4/D/rDcbu5v5VtxezNs5tvF cvPh/vbvP7U34fbmcOw3i3613Qz3t5+Gw+2fvv2vP3zzMjsMxyN+drjBFJvDbD2/v306Hnezu7vD /GlY94e3292wAfH9dr/uj/i6/3C37vc/P+/ezLfrXX9cPi5Xy+OnOyWEu71Ms72/fd5vZpcp3qyX 8/32sH1/HIfMtu/fL+fD5b/riP0/897zyLKdP6+HzfH0xrv9sAIP283habk7XGdb/39nwxKfrpN8 /L1FfFyvrr97keL3fnlZ7st2v/h1xD/D3jhgt9/Oh8MBClqvzstd98vNr9NI89VEv4r6LUR9d373 3TgVhktx+vTK+WH11Xii7bMWf1g+7vv9Wc0AwMjFej777sNmu+8fVwDVizS33wJR/9hu1zcvs92w n0NJ97fAxe3dSHgEF8Bo2f739vjwvN9vnzeLvww9nk2S23Z7vJAhiu37h2N/HDD5YTesVid0z1dD D1ZeZh/2/Rq4vL89PzlNedz385//NnxcjoZxOD1aDO/759Xxp/7x4bjdYdzHHkLw6sLj06fd07A5 Aep/YCpXulH2vIb5U485wdPDrp+DgbzdHPfb1fV3i3FtGWaxh9bOI54W+4enfjeU84sP336znR3G BxdODjcfZ8MvkNOwWB5hprvlYt3/MorNCsjz4+wI6uP2l5vlBrZ6f2vfBrs7/vGtxz+/+Ty+7O6L mfHkZfb1+19m7yHWDST7437Uy/UbFrFc3N++kWfGv3h8EhDmuz4+jx02i9eJLl++mOfzp9dpPht4 diQjL8+HodUf+k/b5+NJXWfKw9lJYeJNvwbUzk8vjufddjGMAHjeL79C86Q1jANOmgdof+dFWzjV /XIxAC6r4eH4aTU0KPxh+Y8hbhbfPx+OS7iyE1r+BQ5+jwFgEW/+K1zwT592Qxv64zOg9W962Qm9 bbXcvVvCOvffbRYw4H/1ZYDMq3oRoRawgJfZ+OFvQOFVDUKUzmibzrIYqa8UIYwKjVOsdZpSpLSV j1EiGT5GaeEcnU1ZFyKlGG9TpRTndOKzeSEDX6kXXnOuvVWeU4KSoVAOgrLWU0pnYnfxeF/IOjph Lr7uK0qyis4WXY38PVl2UxSTJ6STvbNcokX6xmVQ/MRsUpjiOsa1FDZIOttIMZwiVe0cnU3aFico UCnFm8RyHF2p1MZHihBpRLSZcmB0Enw2Y5Tg6zGmRqptaZw0l0DwOQ6ktcXzlcJIMufA2U5RhEgv 9ARvXgVuwdIb0wKVQZDGc66DLYZLNDgp+WydLIoiXibRRWpzMmtlubazCZbzVkRQlzD0hayLU4nP VqWWHAdV2sLX0+BHqRcbU3dP36OEco1qG/lJtlQGI6VRuSkBIFAkKuFD47yBOW4/SilfKHqV0VJT GShjEWoYdpQTOvPZnGier8d51ybGeF+ofpSHFqjvBcXwyKTwGs1nC1IK6t9Up7SnNqc6LQrXaWds 5DjorOH+WkUNS6USTSJ7anMqyY77+JHCY7BKykWOneTzhHSy89fq43PLUkVmyaVTVHWc66LbBHaq MpLjoHooj0qnyZY5EpvtLJWoFoA19RRa6KIpqrRUPlG5aaVTpLjWxgRJedNW+kLxpq0z3MNq6xG2 mAyAHPzlFBc7qgUdRNUUoxr+lfteULzgcuu8n+Atapep5nQ0WXLeoguVj0kW4qYrTV5Gru1sneda yHA7FG9AgeXxVBeogfNWYFhcC9WqylfaTOY5khGqFLpSI5XheIPxGEf1A0qqdKVGusjjnJHAG5Uo KClxilKho9HZaJ15ZgdKldQWjEZVwLk2snLvbwwQz2VgkZHy2ayyhnoxY33hNZNxCipiSDRuUqLo jxSKRBO0FxQhJhipaUQ3HfLbTDnoVNFcCx1Mga80IgjzMajzBKdkGVWlHKDG4DHYVDEl66pEpP7N CiQoVAZWyhYoEq1UJVMcWInckq7HShcCRYhVIvBc2SrkozQujJ5K89m0VZr6EKud5lUohIYUn8na WtkMRZW1kAHVNgra3PhsXuSJlY7ugCLeeiU7roXgK8eb7XTlccEie1JcP4gyga8nQql8TNSR49pG 63kWbRPaHtQa4Y4Ez9ZRzUle5dgqc+JyqyZy72KbrI4jvqniOaqa0bzadUL6Sj0FajYh6GxOas19 Lygdz8UcDJjXwU75jscSp4EqikRnUPBTGTjjukwjhnM2tAmKk5VaifOmdRRV8Igp0BjskO5MvCfo WCkSXSeb4yvtjOHZoIvC8F6Ng7/mXtlFHyvnOuk2gYPkK89dXBaSR3SHamECb9nLjmuhSMXtxxWt O66faoquzPO5JlEecopTjnokCM1Hqh8vlOYVmEcELoa9B6lg5X0+D5PjHhZJANo1dDYwIKnmkOAb 3qH1Y3OSr1Q766h0vHYtU8vySC65h0VR3cmJMYjC1Lt4YyvvV3krFK+3vcUfrh+0VCvFgXc+8w6G 97ZxP+ph2zxX9gG+l0ZnH7TmfVgfXOXZuu9E5VETHhGpIsVBVJbbtkeXYAIH8Ae8MvIRb+J4S7bw XrRPtvG47ZMDFCnXWRbLrSSbyDNvX9Cpp5HWFzdR0/qqlaHeZaRUTmlSSo6dJrsJ+2k2+okxKIOp ZQWBlILGkqBRfDQmt6CV450FUKqm6wEQO+4tUUioSK0RhcREFh2ANt5zCFYLQT1ScE7x3DLA7CPF QfCjV6Qy8D7zSBsCpMMlCiHwSjwEE3iGgvTe8cwuhNHHUt6CL7yzgFK3ZeqrArKdSL1liNJmmg2G JArvUodkJd8RCFkYwaWTReK7IiGryn1vQNzmvYAAy+ZdUFASDqycNn8/7w0GVPW89xSqnqj0QkMS YOlszddKvVgnkGFTGXRSaN45BcXxONdJJXhdAormHU1QkP0zrjuUwdwW0Dz2mvqQzqhcKHZQGHV8 36wzvvL+QYd4yquPzpnqqKxRZlXepcbrC8/wu2i7yrlOpuMRsCsCdSiVW5OCZyhdc0hE2JgopEvU W0ahO16Jg9I0xU6UIlk+G7TNu6BRmsbxNlJ4bwOhvvAuNSiV9+ORBEjew4/GdzwXixZ7AtR+otW1 0BiMoF153yVah3441YJFT5X6eCQoYYJr5G+8OoweOODaRreK52/RI7+lnjwiLkxw3Qls+9L1dDpk GgFjHFuNdEzWlldtMaPnQL1lzN5OIKQ4UbgMCg5HcFlX5SbGVNN4BhkrwjONTLFhL5TaNiiBe8vY 4Cy5dBpqCc41fHyiPjEJaRWVW8KpFp4FJAFvSf1bEt7wcxugeJ7ZJamR/jNtJ/SKuJ0m6Tu+Q5jU uE9KZ1PO870PbNwF3qsBpfAeStKi8m5E0jYUqu1kcKqFai45V3j/IAURee2csCdfqGWhteITtdOE Uw481ifsn/JcLEXR8Z1iUBo/65GixQ4d1ULSU1wn03gdnJKzExwk1IcT7/GZ++uUZeCeImXV+M5D ynaiW5SKnIiAqZiJnlCqPjZqpwk1E+/MpWbFNIXvKqbmMj8FklF7NOr5skBnjKIqo3XLO+gZRz34 ngS2sAXvZGWcFuN7U1lL1K4MOzi64njem42bkHU2XvF+YrZIhLgMrJ7ouyDIeL5XkFG8T/CGkpbb dkY9xzsLuUMOSxECSim02s1IVXmGj2Vqbqc54iQZ13a0lZ9Lw2aj4dltTjhJRuNCBqx4NgiK53kv KIH3BnMeTxNQhGQ7hdHs8oTciu+458sVGQpHYnVInygH2F3gFUtGFsC9PzaMoAg6W8MOB/VvebRt ihCk/pqf2xhbQrwfX+RowoyDIg0OtlAKtqZ45weH0lqm0ikKQKQREBTHI1NR2GyjiC/ol3X8PRrx lOK6aGzTUuwU9HS5FgoOjvC+cnGo6zkHOJ/COwsl6NzRnK+gh8KtvsAfeC43dCB5dTh9Xrl0rk6M SdjA5bJO6OpSb1kQTxOnoCvWuKzRJ+FnZEpVlefXpfrC+6MVO208zlWcDlRUbhXn+XifoirdeI8L h6gmslsUBRM6RUNm4txGtdiBohYMiuMREMddJM9Q0FhHWc3sFJTE/VtFVcJPZ+BQWOXZRkXfX/H3 jIin3hJtesl3bEDxvJKoiEvci1WcleJnQStOZ0iKxNrpLlGbq9E0NUGxlvc2akI/hNZMNemJzA5t H8ljY82y8XM1NSvnqHdBwy7zbnjFrhnvTtbqRUc9eR17gxyJDbGRUhoCOr9XgA2Bid2XptBbp3Jr WijuE5uWSPEZrpvWhlt9wykQfg4FlIn9LBxX85VKB5TGzwU0ADFTW2ho4vDqA002yc9GNGyX824r 3BFOdlIZYFtkQj9BRr4H1jqheJegdVZl/p7ONr673NC35DaHrX/Lba5FhYKXrif6xndfGnpshusn aZyvpbNli9YCpRTVOOKxOScncFCF4SeiWkW5QLONVm3H+/5tzOy4ZTUFjFCucUKFn6JqDU0cPY7B baZxswB3mNaz8a7oeAvu/Gm8GHazPl8qy/36cb/sb96Nt0kxaj173P+clpsr/XHAbdrht5SH58cr 8c2bM+Gw7lerhtuGpwkWy8MOVwhPn1fv+v2H19lOolnP9vQpbhl+P79OPV7IHPZ/xs3L3fkdL/t+ d77mdf2JNOeCZT1bbo4/LNfX54fnx4frqA3ugf6G9LxZ/PXjfpzw7lUoL7Mjrg+frsv90G8+XG9z DZs3f38Yr68N/eEYD8v+/vZ/+zff/ziOxkWx1f5hvHU8vOt3O9yuxO8eP8j729Xyw9NRjsOO+LbA 7ePTl8cP6kJTJxq+jbTTl34+Lha/vnwYf3D+iF9dPrw+09dn+vWZuT4zr8/s9Zl9feauz3D7Gdct cU1vj6upP+Mu4vXj+Pz9drXavgyLv1wf3t9+9egshNMtvPh83ObLPdIfl/PTrb+TiE5XR/8jd0kv l0tXp8uZn11aHa+0jrdWd589vVn0R6j3ZGrXm6nnwScz+mIhL7PFMF8C8g+f1o+vF2rfnmWyWh6O D8MOd2+P2z2kebq6+ccTTZrZYjv/DvcF8en03MSK5uE5psBerzfcv/0/AAAA//8DAFBLAwQUAAYA CAAAACEAT5dJ4HwYAADMuwAADwAAAHdvcmQvc3R5bGVzLnhtbOxdW3PbuJJ+36r9Dyq97G5tZay7 LNfxnJIcO85UxuOJlZlTeaMlOuJEEnVIKk7m12+jcSGICwmQtCc+Z2ceYgEk+oKvGw2gAf7t7193 286XMEmjeH/e7f/Q63bC/SpeR/tP590Py6tXp91OmgX7dbCN9+F591uYdv/+43/+x98ez9Ls2zZM O9DAPj3brc67myw7nJ2cpKtNuAvSH+JDuIfKhzjZBRn8TD6d7ILk8/HwahXvDkEW3UfbKPt2Muj1 Jl3WTOLSSvzwEK3C1/HquAv3Gb5/koRbaDHep5vokPLWHl1ae4yT9SGJV2GagtC7LW1vF0R70Ux/ pDW0i1ZJnMYP2Q8gzAnl6IQ0Ba/3e/jXbtvt7FZnbz/t4yS434LyHvuj7o+guXW8eh0+BMdtlpKf yW3CfrJf+M9VvM/SzuNZkK6iaAkqhQZ2EbR1Pd+nURdqwiDN5mkUyJWXrIzUb8iDcqV4c5VmUoOL aB11TwjR9E947UuwPe8OBrzkgjBRKNsG+0+8LNy/erOQmTnvQtGHO1J0D+2ed4Pk1d2cNHaCkvJ/ JYkPBfnhF7JyCFbQH9BM8JCFgAuACWl0GxEYDqYAGfrj/ZGoNjhmMSOCDQAxuVn4qSgd4ALguaMY htrw4V28+hyu7zKoOO8iLSj88PY2ieIEcHrenc0ITSi8C3fRdbReh8RkWNmH/SZah79vwv2HNFzn 5b9eIf5Zi6v4uM8o+0TIbbq+/LoKDwS50PQ+IJ18Q14A7EB3SHSQoWOUc0MLFKpY+E9Osk/70Ehl EwbEyDvI/3MQGhCJNAG8eB02b2LUvIlx8yYQvM10MW3OBbj2pj2CFlFPkGi/Dr9S+BUbQE9kxCx9 xQAk+ytZvNJpDGclhkHe0EhUvqFBs/INAxLLxTDgrvwFA8rKXzBgqvwFA4LKXzDgxf7CQxxn+zgL O1n4NdOxan+R+LYw8XmDkPJ7YxWg11aJDLGXnLyaEC8JH8IEoq5QbaxXAtN3UeqlE/J8Z3HcbkP/ 126Ou3s/7SA1zYqq5OloVlT5hpcVIVdeZiRpTXcKlcxRddeQir3oLxx70V9G2sV+Mi6jbIuYLQ0f +g0DFRYodm6DJPiUBIdNh0TGPqZyd7zP3FjFYOhYO6a6y5J4/4mwVqoRiK3JwFubzOXusAnSKK0m 1FD1N3GHGICPrpdkptN5k0TrSu7GdDDQ1GD367fbYBVu4i04986Sjwke74M8d3RaUclcQyS8iz5t ss7dBmPsSmITSz/ZNUHb551TiraJRZSqxp36cGKBsr3xn8N1dNxx1ThMPyZ0UPDoZoUEsliuIupo /Ung8FA9g5pQf1yzfQf+aaTn3z7pY5cuoIFhzfYd+KdxZM32ER/l/evtaV7DQpXwfaVtT71t9yLe xsnDccttoNI9TL0tWJBw8hBTbyMW7Ts5iam3BRfcZ2e+WsFSjQtOvfsi96MeVLy7g1JBY3OXxbtT FLfX95DIu4MUWgMPWs18rQchb6f7PvwSkXVwYpB6dGYf0nAUEOFppTkPLRoAqk7zxl+PMFetpDKw +DxXKm/3sD6ahh03akOL5blSY3hCTfqAqdnA5wGmZiOgB6FmQ6EHIQs+7DAXY6I7keaDowctb7cs RjGEnbNnbjhcekjk7ZaN46ZD/GWxXjsW9HHTgYp3B+njpgOV1sZNB1reHWQbNx1oWUYNex/JPlUE Aw6EvMdNmZBw3g6E2nHeDoTacd4OhNpx3g6EmjvvaiLtOW8HWt6+wei8HQjhIz7RnSAkx+8OhLx9 g9F5YyulE9Ami0d8LHKg4t1BuvN2oOLdO4pDFX7OgRY+4oMEhZZwdQ602nHeDoTacd4OhNpx3g6E 2nHeDoTacd4OhJo772oi7TlvB1revkH4VDnydiDk7R4EIdl5OxDCR3x8g9F5o9U/ufN2oOLdQbrz dqDi3TuKQxXO24GWdwcptITzdqCFj/iAgdFCcPsI1Y7zdpCoHeftQKgd5+1AqB3n7UCoufOuJtKe 83ag5e0bhE+VnbcDIW/3IAjJztuBkLdvMDpvNMYnd94OVLw7SHfeDlS8e0dxqMLPOdDy7iCFlnDe DrQQL42dtwMhfKQuIR+J2nHeDhK147wdCLXjvB0INXfe1UTac94OtLx9g/CpsvN2IOTtHgQh2Xk7 EPL2DUbnjTby5M7bgYp3B+nO24GKd+8oDlU4bwda3h2k0BKuzoFWO87bgRACs7HzdiCEj9QghFbk 003tOG8Hidpx3g6EmjvvaiLtOW8HWt6+QfhU2Xk7EPJ2D4KQ7LwdCHn7BpKaC/mizhmtfQsIXPMM eFaDM8GBpZNcCTIB38sp/qXj0rAhQS6hB0ULPFxFXMTx545bLvjQAhBnUtH9NooxC/wbZulI5+GG 05LTEstfLjrX9MSb9h5Cqph5A8cF5fOBeB6RnBQEPrNvBzijd+DJ6KQ1OBFIznKyM394LPYtnABk 5/jIy+RgHzyIpyhZMe7bMqr4NxzBXfNner3hYtqfn1KJ2InIGI7pPmzjx9vjfpWJJ+kj5PgjSW0O X19aa27UmvUfxzR7TzKV3+5z0rTBVD1+iSkC7PDlSPx4Xzh8CdqBrOz5Nvq0JydzOb37IA3Jm6Rl ei6T6AROveI/hXOu5915EtHDj3h6Vfq9SvkPVJd0VJWdw0j/lI6q8rMZ9KgqPyl73v0jePXTLWME OQCGsMcq+lj0KkNRHw/yyv2aH6hE/vagCa4AuctBEZ95OW/rYhMkVO12QPQuRheLCX2KAeJzGB5u gAzSIz/egZJT/HVYrFG7WXzgxMjpZnZc40+AKho96WVAM1rliqSlFg/Sskb+IuDdh3B0HLgbULix Y8D9UwE+UUXApSIRzgqC5Nvwgdhln+bcbQANoAVWgIqqBOzjWXzMCIF3X7ZclQgudyTn6FtGOzgx fxM+dt7HuwATDU0wVx9DTnPADxkMdMADT01APbCCmp0GJ5a8/oUcVUZXxvFLUemB+EFNxBtgvYdr CTAZEkCNyGCYxr/tmM79G4cZARYqWu9vPEet9bfUIaiehsofWpU/pHyZlc96xkP5w5rKN6hsYFUZ 1whzVczZ5yob4NjWUGX0NgWTEx6VqYzp00Nlo5oqKzihUR+nSrkTIgUWxCGLJYgboIAN1Te2Im5c pj6mWw/1jdtQ37SHkXGuPlJgUR/XjhV9HJ5NvOXEqj7mn1WDZcUeipvUVJzuwrBHOaDg3zoxz9Qq MAu9mws8bU1gVHZDgU+tAjPDVQWmMwwGS49+Pq0ptuxfcMB7iJI0I2EgGQwtxoGd1VAzM6tm2NUK Zs0wtXloZlZTM7oF8AEHfYLdAlYQiwcrdhmAZWbHjiSLIx94IFmd51nOLWOn5Bde8FCKYSZfGqDP FY7EQZGd74wcvC2ZjeLB3NIpaQcfoajRGYTLb5ClKg5hDni/pVM7+INOLOEeJkQnnTuvvwa0Kai/ CLfbn4OE6I5OVSyPkjgeVAW1/R6OokpT93GWxTv6hIlUgudxrQ2AWmVm6E8ihF3fe7wTAWYTJTq/ iclCnWYq/Fi1BQqumrbzVlidEDNXIy9YiIwoFivPWYtLE4aZqOyHxpNimDM4LYzGjlyvYGki3t2R RRfocwzwyd9kXWUBNwqoxrZg9qMIkUvnJcJ0WBRhNG5dBIME/0utQhGBSIv9UykBIJLaHfzxlkx+ cY6GigGls1r4q8IB2sGjTw+JfuHOCBN+aqr+dMwZ9mGzDC0wR9HQYmeZ1VQqW0Z8vz8rhqbt40VX /WJgwQufk1WKYMYLj4xbwIs+o0W8mKezpKqe8uHKHQCgJ7DLEAOHGTXE2JlmNZXqLiBmMsIdRzGZ aR8xuvIXQwtiQN4mHobPVQWeIBTBew3hX+6FyPhNnPghhlVcXI6jXUYf9ey7woimrwEgyNgkVXGm pKpWf7GpuiejZSDT+V7YeWY1Xhgb9GZtjGIFZesrBqhs83IBqarDeH/a406orTFA53th55nV+Cl7 PMYF9Sc06He6U3pnHHm9IjdhtPBHHjRgtwHYRW2VSefDdi2LLjOUG13um6eSm48kTO72zP3mF00K KDLFTV69V9z0KYwwfXoTnQBk3kMtGdXlOsriJL0hlxGowyWt+y/Y6iC1RjGZ9HYrY0vGuITPh5Gr qx78RxqErvFZuysB2OWvOvu/mnmmF7OiOLYlDLs8qPbHs2KXGeyK3HVCL9vDwXI05ts98CzXA05n xXg668GRdKqTovWJTdZ9fJvE8UO7evuHrrd/VOrNUz8FSJNxATQgIN1vfVS+/F2X6XejTJdMVJs8 gE8cOYll4B/AuCF1oASXV7rDhyIEnxLVeLuMG/Omsb7BA2dLkeAfK448hk6KJAZpZqn3/vASwQXZ SQTQq36E7KlDMfIAe7CFjfMwyZcGbb3AJruw7hOt48cLuDcwiYu7uLC9+++VVtHOZjTpawgPhFsB kIurw/muVp3t6cLqa4lxULwgAFTM0KpODg7FWoyLtxaEUeTZ0cWQn9hyWL5vZZtXeq7oVbiqWtkN uagoRaPMeDUd0pbyjrDrkTkSJzcTAZ3q0d8VSBKPqsS0qimQaCvfE5CIAp/cdoVnJ/KTS5zzfERV zw/8mudEJEmaUGa3W2VLwg4zaaDCpYiI3buPC6UiqYYPdX2+hegTaGpy4z2lVpFJ+hHFhmJT8pgu j3xcnaTZ3LKcNWAMQuUgiyxEyTGWsjAlIkrJ3TuqycciCWCEhKryuArwCtg2zJNTox3hrEuL81eT t0CbLsOA4TWS8agWo2Woyiejr5r7WO0iC+6/ZKy9utXiMigymagMWpsV2gNjV4CQ4Qb23foIfxUc rLLTbwoMTuX78Nx5j/PZYJ0e99QwzdKzaXjQkoYrF/0tdiYShV2sSzzsYFNDxtEzaJim4tk0PGxJ w5Ur3s+tYZr8V89reWKYZu7ZNDxqScOVy9zPrmHG0TNgmCb32TQ8bknDlYvyz6Xhpxtnr/WEQigy jbNsbGI6qb0UKcd9/dlpcUcSC5C4nkqFqUAwkBfXf/Ixim5BV0cgnrZMsyFtSJu0hDSmclv8IiLg J570q9p8Fm9J0y9tGp62pGGWLfpvqWGa6WnT8GlLGmZ5lt+Php8xaqUZozYNz1rSMMuG/NfQcGEm KFYw3pIPtOnn3GDUwO+2mQYmeQLoPI82rEnwscU+WSxhWc+VoiyzsF5ZaGFSojS+LEO7MLunZ8xa T3N891qbdkMR8mlHnaRLyybPd7PdIR3ko2f3wq+Qd00FzGB1q9kx0ov4mETwmSA4a0fWs1zmiMor 1pmiedeDp93VibjNYCb5M/R7XjoQoIp/zM1kh+TVcqj4WlbOjG5epI5xU5IMykRxA7BsWfUyF4Qb yznXU+JkztksWXEP+duV6xTM9mTWIdONp5RgbX2t6+liMu9stmflvXIFwMR7vd11g9717CuZdzaL sfLOZKv0eQW910shK/B+Qz96qAYSyDqrez7TowQtpkcrS0xPZtdLj/VMr2Qt+eZK82VQZNIjT4+q 5Fc6Bc/DBjFHyydRdFMaHsB1Akc7LBNEzxS5MWeKVAnSnJdbPVMEitwcrSEBqZCrO6SnNyB5Jc8+ KjwwheMeSMr2QL8PqZHlT5hSnApE+jM+vNqoDIY95iisT0xOKzgdsjwqu7DD0ZgtD9iogL5Yzozt idGAHt61UxlN4IxTqcZGs1mFTsdD+EJ6aRvjKc/usXE66Y8qdDoZ86VcaxszuGmilI/pYFbBKUCs QqenvUkFp6ejUYVOT0/5wr9NltlAbAnDI5ji/t1E0+ySDHJRhXoTxgsMpqme6y15FYLpEg++nOtu k5QhWJVoRJ7X+gxIxXTPZxyelvMLbaAlZSbhhNCVkpXl5zQfxpbzawPP1xaemSx2ntmGwD2IDLy1 NOwv5z8ZePzJyGM7oHk8y7UOp1w3hNSzAknPv1/OzQn47kCSY3USY8oZJ63HnMv5e0OfvTf2mbsI ea/QZGypW1owBT1MXprDZJ4SbLcDZUkIVI3pTmxE1XN+8fg0u2yJjCrtCmYwcbOFc8Hk3S1eZhPW UfFimge3wunrm1m8gk+NgtTwEUhxvRzX2rBy3VdSN7mHgV7HVZJ+zCMJKegQ2f1bTH8+767j/PDc bEJ5AFn5q39RECIuToLYHPTDIIN/kzDE+Xau8QTvKkWpz7ujQXEfktRiZzQPY9RcKmDaZV3Q8Jrj 2mDDvWIZp/rUn+CUhdZKtIKwbgpgk2PANj18t2Ee7ro1XIjiZEXoi3hEEebFO1AE01ADS5bltV4k x0fkGh5IX9kjArHpjN6zTNImWxV4bgbA3/Qgjdwt+hofkcK8tgfdwsRrJIXtQqgGfaFnYBAp2PRU 74uKBIxG0tnyMRpIp9/fRKRj7lWXjoktD8BySNtEusFAvSSEFIB7aSCdflkTkQ5bfTzTpWNDvK8U ulsU9wbKjmIwmeGlmsLEsEASULjGe1roMz+R7U6/iIlIzaIUXWqmDl+pZdHa8h5lU/OlHrAvaRSg iFTM0rVB1RYvigymihQnR69eItHHuSYRFOF4WsncQwK3B98m4LDx+qNBDy/i3px3pzOaqy9vn5Kn kuBw3oWk9kUIVwrjnVBf5vvVhtzuegg+YQHcOkoLdkECV6KS+O0bXszAXFpJwMrv4BRXLuF0ouR2 WZze2W/i/IuiV69w1TDZg/C78Sa1yEp2CUPFw47BJ3Qp9gxfj268MV0G74UO70VTeJOMChBi0x68 Z6dsTCuB90uAI3XHbMLEV+BfBkbLDp/xPZ4nRaqeVfPxtT9S67tY2HZhwe8LB+ELcX98U+dJQfVG d39vngtUXzFR6rwrVo1gIIcNs+qB/CV4upfh1NrPCHPNjP/45u5nA/agsDS2bCGg1RdyP7KFXJ+A tp4f5ZDPr0r5V8L8C3GszzJa6xPBj2wi2AxlpvkP86P5DScv9dqQ7B2ZT1AP8DIc6D16qzxbCz7X wUoaLqGXzVmWhlAQypCwsszw0X8yw6Yt5MM8ZBZjmJSDzzqlqVF81UssCkV88sZ36cTlKmJDzmfB qEwLH/Th4wPVgY+FkXuZ2cKEQVLH5YfJAO7dRfWXhMZ8+cHtyzb/v/ZAMx8AWU97R037IVBhB6gM wb/pCP7NbMU1gK0HKGSdbAL/ExLccB3jKbFwuwi22zjeL8ntH2qCMavD2y2MYsiL8PLVIFKjF+KD UvrGl+v14gCZwhfJBpPp5IrNKlhujYqpZbCBzxrhos18n0Zwe4UowEUr+ot5du7iTLleIs+uwsu5 hsmqasp03vSAmESLdp9zFzTQt8sC4rN2hoD6NXwvLyEw1YCe15iGXPt1P3Z9FhXYm096cM0yNs4A W7hMklwleXXF+oijka1jU9uuQF/BQwmByQGBW/F5QBVpeHwgrzZJLtu3LisPjUq+FCfJzbJRClN9 pCl9MFD5fUN/Fz4OiEX51hfPDoKrOEFvLNkD8o3ZDzUBVd49mvJpwwou4wP3dwy2d/RTg0jDOPfB GAqqubcVYVLR/yw34Q6yTnbRPk6uif8h/HDDMFbit9eMNas0k1pbROuI4iQPUQcss6CwrsTKtvSr g7j8Hu5ffbgjnNxDI+fdTfjqLeaYgjC14EXR8Xt4r9kTren8N9T9D+VWiWLLgVV3XIAk8cs5M7On gZtINKJ4oz/nEFgzRMKATBTMUNmnT+Ev5aHGEDvv/nzXuX0DaZjRSkaXWo7AUgvJECi9Tm1KBPb8 UEABT2wapOOJWkMFgAqjo/BPF/GOfJzzvbhsTXVRwR72DgO8Gi0RDyG3Cp7quGhl57n4JWwkUnTi MrqYE89t0HTTZK2DPqpyjIGZpBfipipNTI7NmNJJuy6xmYda7KFnoftLAmiVNxUPrL6V69YkWmzg 1VIsuR+qxEHR++tXpUne31jJjNQnvS+HHh/GCubKk2Ur7NIcNzDVwAej/wjhHLTaCxL6UvaIySY1 Bcs5EVqlAaKMvgtKW7DleyrDBUmTrR4RPREti6Kqk6miw56xB/2SznKd2PVWhWpFZ98jyOU+IUdw wI4wkmkCedeeu98ewyyCT8Jr+M9rTLC3D0VFDY/nw7mYuqPXJ8DAP0BMekE3pFWz7yTzL04tScmb JFprXGFNB6tMbGF1edg1NmaUX17CScNLCiYWVGnhtxx8w7Ig8E8i8WLwnYfMPl9lW8QJXLtNv+Fi +4A05mYQnvBTg0BcyfCh92Jw7BTTg6rfLUsuqn4bt0Lrko5g5rAOr5u9/lu916G7oJ9k9b/kT+RZ xrngQGJLzZRWrNxkR7IJyQOaXG6YKrNZiSGass+eLyeTxZwtQ9ssbyC+QcGeYJFp0XMWFhtGV6f9 xWtipDixvGAfW1+R3Saeiio8rU9cy+O/9Mf/EwAAAP//AwBQSwMEFAAGAAgAAAAhABlcQqfDEwAA QPAAABIAAAB3b3JkL251bWJlcmluZy54bWzsXc2O48YRvgfIOwwELBAfdofNfw48NkiNFGzgGEHs wGeNhrsjWH+QNDveW+BHySWP4PfJC/gVUuwmqSKbbHY3SYm2uYfVSPzt6qrqr377y69/2qxvPsWH 42q3vZ+Qd8bkJt4ud0+r7cf7yb++n7/1JzfH02L7tFjvtvH95HN8nHz91Z//9OXr3fZl8xgf4MQb uMf2ePe6X95Pnk+n/d3t7XH5HG8Wx3eb1fKwO+4+nN4td5vb3YcPq2V8+7o7PN2aBjHoX/vDbhkf j3Cf6WL7aXGcpLfb7OTutlkssxubhuHfbharbX4P/o12+3gL7/thd9gsTsd3u8NHuOLw48v+Lbzh fnFaPa7Wq9NneD/DzW/z6X7yctjepaN6m48queYOXuDu02adnQyvXX8uo8Ad+8iuOHADrXhJdsnD bvmyibcn+nq3h3gNL7zbHp9X+zPddO8G9HjOXkk4YDTY1z2xuefl5JGZ9IfD4hXmPnvw6567XQUx nthFmzWjQ8JQZzYq35EYEjOS3CJ/B5lXKD4zexPMfK96pDlz0useZLCNQP31sHvZ56Par9rd7f32 x/xeiSpQeDPDpaKOh3ZUugGnK757Xuzjyc1meff+43Z3WDyu4Y2A4jcJR06+AvW0eDyeDovl6duX zU3h2/un+4lBT9keV09w7NNifT+Z03/efHKbXLx5WZ9W38Sf4vX3n/dxdk6ipNYx/Zmddtrs19lB MrNmoRuZ7Mj6U3JgBR/Zw0CLHk75yewsUKHzTf7jU7xcbRbr/Abfxz/lx96Qd/nvf1tmt1nHH07s 5/0/Dsl7n2DU6Wd2DjxjAn/vd0Bx17aS02/PJ662CQWS+7Cj8OV5sf0IA72fWK6Rnk3vDlfBcOjV iLaNpCY1pPbpzbVI7RPHD2wrzCkCb61G6seX9TpOSQeDwpT+9edf8vvqU5rYAeWEGlLTwwJav94d 2IQe5rvt6QhnLo7LFQjwd583jztYa2CaQmDfwg+rLUziU/xhAcybThy9i+bEmXUTRxh9NGTE8j0/ mj6kM68jI/1PnGkEKfEyYcIyQpLDA584q27iUt2kMXGmMZsFhAx54gLTFcxbcnTg08aWLX5N8qnO 1lKUURTNZjNjyNPWakkagpp06qTN11aTUTCd2hHYHxSL6KjJ/qHEGRxUqcnkqEDcNFckt47UVGNr SYg7AygR+kOGEq0oPQQJ8bhpMwJiEncmVGzPnx8Pq6e/J0B8nQBxJgwYcXuGNTVcs1K7AfOd9msw /w3bCAzDYCi0GYO3xxcYLXjwcgIx6GVyKEqHhzJATJLnFyjB0EErSuxyxSQAyJgOxLb1CDHdvRxW 8eHm2/iVEpJh3tKvS7BsSj89N2HhIpVMjkoO/QU8LeBg+QSmZSf8819VuoGQ6NHtBzDgEh8eeIVy S6H4mxqBLI5AjLE6JpCU5YUZy/TBjLyuhNkcaRixOiWNssTBoqFHmLIg9SRxTL6whh6GxNmWpsou ShejWvE3NYmj5kpBcQ9D4hxYUK8scd4gJc7xNHX1hSQOAjklJDAMiXNtTRVelC4diQMzRMWT6fM4 NvIJMQOPrevVfpVmHGuBtefP7dReLJp7+Zy5bhgaUcScBM3orQMDEK+zDfZcLZANDytwaJ8xSP49 wWz5F0W0xmFawjSC+oq7WL3Euzn4gn9YPUHwKzO3sTv4L2/ML/IDknDXL6lIm9kowG4FR3oDBiW6 Cj+d+9l2ud4d46fp6rBcx/kY8ODepBYY8Jzk0IhZ0nJSY+Ph45xqI/UZE/K1LRUjwXxN3JLykRoO B/n6ZUBHmQFNUkJ/UsPiMNklOPDs5JLkQNMpITSpsfFwqhcO9JQ50Ax0lAUHgfrlQF+ZA63yiiE1 SxxGuQQHps5DeR1oeXJKQxFeBDy8mAdT4j4I423N8IJMZzMzsu18BYDlOIuW5vDCDuzQcF1ZeCFw k/3v3//JnyQQ6cxrDJ/wEjQEjyPVieesPlKt61c7Iw4GF8/f1YwzhEACywwtcNZrrme9E5K63uop qe2ZO5OuHSmR6y0IbAC5LlPV6tCgd1JSb1w9KbWddV2REqGsMDKnZmQx4g6RlInfTkBKXbdeV6RE CA/CQ9PZHPJ6El05QFJST189KbUdgV2REqFKKwrM0CGDFXDq/KsnpbZvsCtSIhA7C0LDDn3GpwPk SuomrCelthexK1IiAG0Z3tSZBuFQBZw6FgWk1PU7dkVKBNqJY8ws25kOlZTU11hPSm1XpD4pFW0E yMBKEi5xkhAxpsSZz1r6IIlnuQ8PDT5Ie2a7ZpTm1zb7IDEg2393+rzOE2mjs98Ju6N+/VkqQtpg OhCX2TpA2uzMossHDsNanqe52g59m8w7V+vLbO9vLjr8kBGRJikwANepNpcKHGZkgs8qW6wvE0Iq 8KHoG0bmRErU60U5GsjalznRNZ8isyJPpukcC3eTdU3TAer1+5gtkCwecqUCfZkVfQg9MjGGLvR9 mRhdCz0yNYYu9H2ZGm1KLYorPTI2UmIOdKXvy9ToQ+iR2TF0oe/L7Ggv9KrmB1/MRTwnJHNLmOze HKMw5nPTDIg4BcLwaSrvnJpAzeaHMFQsV06H7QYTjNuC3ZBEuBjWEET1Mcinr09o+ZCGI3G9e40P 38SnU3zI4yvYeHpjqscefU9jSBhisyFFmvY+HdI/d5vFtnpEVtWIDquPz8WiRzxLlhMUh0RyR7Ng ljDEzZhMD+IKmU49P8G2wOLHxqoU0yHvNZOZ/pjOqZoiriwVz5BjWBpDwhCvX6Zzq0YkZjrHK6kG KabDEKtHplNPSXBtHbWAQQ6bod40nV81RUKm80hJLUjJEYYY/TJdUDUiMdN5QUk11DCd6hLPl/2S 0Aug0iatkdLNcnQdw/Ts6SxX9aDWuDSE6y7xhPh0cctdgyUu4V2Dit4pDg0YnYAZ9ZUfxKHfkXIg wbieHw6vPqZrKg48Lbxvj7WLNiEPORhzdOr9lfSqYQJZlq1HoO7MZQ6+GNcylzFhbCg1FEExXjmk nCNl/LavYeMQ0kAkzvZUlWpPEsfjrWFInAMLoxZjdSdxHHYbhMSBqtYjzIUkjoOHA5E4N9BU4e3X OFWwybcqIfPIMabelMFEXbDpWxbkvzWEs40HqLsxIBqcREUu7k9KsKXAsueXkxZYk9bAk0hyoE1u mUq4KTZWGipi6gYbL46n8LhaJMwArFXoZlSEU1RR4HLJS2SOn7MYBInOGEOM1TPQDg3DO8aYbPlR R75S5Vtj9QzmwLF6BjgQw50LcOBYPYM5sKfqmTQrDWfGmabvmvMore/RhRJmNHfnDkkBSU117rD9 VrB0wnsnKAe/PueM6i9KUIkYhA7bJgdV9ZAwCujXYasRmuJcTzUO2+Is8X6ieQIbNQKIHYemmhxF 1TOEF/+eowTqoakmF0/1kDjHCwBtvRlqAt4aoSnO+yLFdLyrpBemUw9NNflKqmcIr/c9M516aKrJ y1E9JM730BvTaYSmOAdEDdPB0FT6bxC+1aJpQ/Gl5T+08xYEgWtOp9O0PSpeI6mJjs3KsZFc3lJv CE557fJVKRehoscFow/WdXAgMa+xi9xECDbHvHCUF16w1XrtIqcoXhhnDUm8tMtC27vbi6YCRm2M QAMJcEF+rpqru4Nm90XSYATISDOE1avflnGK4oUR5ZDES7sos714qeJTvj+16UShF2bOI10X1DQi PrFn51ZJValTgWXM5rYjG81qsjXVE6TLTVnOLVzSiqPylhaKDIr9VZRB++15NLZ9Q1vDjIGrMXAl HSGXaTx41mWSodMxcDUGrjrlwKG0fYNVLLG+CoGrYB54/qxl4Mp1gxBy786SVokaPMOeQ+MHydQQ QUuHtEsdeM9wVZJkBmpDqbxn0fYWgMeyE7GVmhyFseVZ27INHVTy517vlrv17gCPoa3qFi8n2lkB 3kiQlMIhlqul1WVkg89sCLjb3tjhgWPbLjqRjB0e0A6G3WiCscMD1T9dtHWBBMSsgDVTD1ir0sNY rTYWw6jkuyvafkPy/WXEqtGlY4eHzoV+7PDQmdCPHR56Wel/Px0e+E0uLGcKSRPQGhqSsOv22Gvu 8DALprALGzSspHepibFb4MH0o5A1uZO29tgtSx3mvs0fhM2RN+SL/HeB46FBxXdsjiiuhZxVce28 vQZqtTQuwMQCbiknL3Kx9t7yXyQz/RqIoGoK1GTMFKNaODeQhWx6SdOSa1vRRIB2HZ+ruQBnE7KQ QG89B+SyCRuI0BJyVxOBw8a9iYJk/mEDEVQBspQo8LHvXkRBLmOxgQAt4Ww1F3AR7v5WBbkcxyYi JG2LWVlXdia2PptC49VE4GLYvYmCZFZkNjT4rPK7qcLGGlEAYijlUfI7jVhBEJlG1pxDN07twMYo pgO7DogwXt5OMFnSmzEe9jhz0P2X/EkCMIfZqhyjlvVpqLiKYTYUfcKtKSHlDsJ06Ddvsn0zAw7e DaS0WnuPj/bJKE3Qbxi5XmMqZSLMacILlrh+3antJY5DkQORuDG78l6cvKy9lUZ3axqHPa8W58QS 1wQh02UaPua77ekI+EwlgNFe4jiwOhCJ++0kXMIecYm2xakTtkMi52FGTQt9Z6XpBaZvN6ROXBXI EsMrJj38YZGsB9sv9xambC/lg0WyDkufzbNmZBnoj4JkLbOUViRLoO7WVc69OYh11QqcQUvcYJGs q6myu5Y43lc6DNvRsTRVeXcSN0wk6xJNXS1V5dp+jRsskvU1VXh7iVN0yZr8xgr2jMBWa37LzZ+D h9C2TSLVdZm9w6VdsupNlzuvDsQRddofAtbZK/hklXsyX9ZexUiWUYmttOp924Q+fanEW2znc311 ZIFaeymv98kyAg1jXW3q1FPrCOluXcVINpMwMNo0uhgJeEc5CtLU76eWMBdaVzGSHZLEcU2FriVx GMkOSeKa2hTVMlZ3EoeR7HAkrqnZUS1hLiRxGMkOSeK4jkqXkzhVJMvvH2LPzQdrGrXcP2TmeYbr WuI+jFPfDY3IDyXhGy2Cx3tqlZJIv8nzC8YkUig/TW0RqBybb05ZPgtHw1KWhuS+ZA2ZMmMS6fF+ MiaRjkmk9xNQ/aLMOXpYEBABfV6RT42hJsukjZKoigZAb2osMiaRjkmk95M3YxLp5M1vOYnU5Lfu gG0QiTc3WjbjtIyH0ItCsccyNKa+PycMazZ7LIWtj9U7HSXgW6Bha42Y8LBarOmVIeQs3E/y74kL Pv/SuhyImX7qbjmp7RTGtkhjW6Si2xP79i7Ql2vcz6Pg/HYgCQrrIhu+s7IKwX7H2G/WY+GYHNAp DCcotYqUGg72dV2AA8f9PPCU9bSfh8laEuHcPieMiGc+tMzte/Ddmd9UiJzm9nWyO6s6vmioUqm2 4HAQkwlBb7WQGtt52CW5Lvkvq4eEI45sSH0ZpZJFvpjvuVKQmuKs4mI5qJLdwnAS1wpeSKRmiF/7 e2M6uQJcPKTfiScED4mrhpBiusEu9021C9VqgV/ue2M6OS8FnqGmqoPqIeG4U7+aTtLngIfEFQTU MB0MTaUM1eS383BngeF4cIAWkOqWoVpT2NcTGonkwRtQa9xO864bQqQokm18OHoQZIp1Rw9CznOS BcljY+WxsbK0/3JsrEz7oIn3chk9COt8veNhhW5GYrr+yXDgUBorQ/leuTrQc3wPdpakzYT1qwPD 0A3NpEvZ2ObiQPflrognc96IQZQu9dv2QjFww3k3BlL/yzk2Sp6A2vBWn/nU1CiCzfdAtWnkJAiS hiXbEmN7aOxxcZUeF4rixWW0DES8OBfOtcSL9woNQ7yanEK1+qef5GmmeZjfVj2mLtA8yuUKTa6l WsJIJU8rihcHLQciXpyz6nLiper/4jf+8FzXgW67aY6zrv/LnoeuMYPkGCE+zeolWhe6aayg0L1C GGyo5eTuRBwDVEoK6JcoR4qOEp4xoiBeIKRHtdMYA0g2hCGFxxxSHFKN07g+PMYYdK4J+YQOW7mO tniGoBtEcTgltVI9Qzg8ljGZHoJtZDqN8FjgaAwJw6p+mU4yURjPku2W9IoU02Eg1CPTqWfDQEME jRnC4bGemU49PAatDDSGhMFGv0ynEx7zS6qhhulU4QHfid8LiONOYYvPVuGxWTifQpWUuJDKd1x7 ZrpTyTURw9xSDdUZhuAaqrc5OBEESppqgkhAK5KAsNmZWBmQ5DBo27zDUbI5KlAOTu9pQ9Kr+biy 4cMnDJg6ycftvTBPlmrSfv1ZqstAA1lVy7FKkKEW5P6e/WhNJG1X4FVL0u7sBgzprusaaCAlcFuW CZydibUjPYy1oyx39uFMwKCSEZX90qm/pRuhV929QJasXQs9BrXXdp5n/AefVctTy/0QLiD0GFIP Wuipi5BhnIzoWOhHDyIDR1wGQUasGg6lTsV6sg7K5wjAD4QM/n//lFTsJz4tlJX3/imTQGgpx0YE pyYnFa5jhXWV15k0rw5AdNV1rCiw8jqw++qfx9LwKq8zqQFR8zyWQlh5HRG9J0sNqLyO7pdc8zjm sa28LDMtqqjCLLnKyywBUZhDsvIyEU2A0PWTLngcETALzcisIUpahVn5nqI5JwJmMUXMSQTcAoOo 5zIiYhfRBBIBv0DrPsETBRzDHPR1RBXwjJA2AqYBD3f9m6a9tCtnMS8pq2LutHVh5YXCB4p0jIjf 0srjygfCMcEQRXwjEvs0U7nyicDFgicK+EaknoD9a0UYCqMEDxSwDXC/4EIB37A+6Bmjss/H+LDa fvzq/wIAAAD//wMAUEsDBBQABgAIAAAAIQB0Pzl6wgAAACgBAAAeAAgBY3VzdG9tWG1sL19yZWxz L2l0ZW0xLnhtbC5yZWxzIKIEASigAAEAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAhM/BigIx DAbgu+A7lNydzngQkel4WRa8ibjgtXQyM8VpU5oo+vYWTyss7DEJ+f6k3T/CrO6Y2VM00FQ1KIyO eh9HAz/n79UWFIuNvZ0pooEnMuy75aI94WylLPHkE6uiRDYwiaSd1uwmDJYrShjLZKAcrJQyjzpZ d7Uj6nVdb3T+bUD3YapDbyAf+gbU+ZlK8v82DYN3+EXuFjDKHxHa3VgoXMJ8zJS4yDaPKAa8YHi3 mqrcC7pr9cd/3QsAAP//AwBQSwMEFAAGAAgAAAAhAMB9NMngAAAAVQEAABgAKABjdXN0b21YbWwv aXRlbVByb3BzMS54bWwgoiQAKKAgAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAnJDB asMwDIbvg72D0T111pasK3FKV9vQ69igV9dxEkNsBdsZG2PvPoeduuNO4pOQvh/Vhw83kncTokXP 4GFVAjFeY2t9z+DtVRY7IDEp36oRvWHgEQ7N/V3dxn2rkooJgzkn40hu2FzPnMGXqDjnslwXUghZ nDZbUTxvyqeMRyF3j9vqeOLfQLLa5zORwZDStKc06sE4FVc4GZ+HHQanUsbQU+w6qw1HPTvjE12X ZUX1nPXu4kZoljy/2y+mi7e4RJuD/a/laq+jxT6oafgE2tT0j2rhm1c0PwAAAP//AwBQSwMEFAAG AAgAAAAhAKnIXKqMAAAA2gAAABMAKABjdXN0b21YbWwvaXRlbTEueG1sIKIkACigIAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAALJJsgrOLy1KTi1WCE7NSU0uSU0JLqnMSbVVinEMcNSL CPZRUgAL+CXmAgWBYkoKFbk5ecVWSbZKGSUlBVb6+sXJGam5icV6+QWpeUC5tPyi3MQSILcoXT8/ LS0zOdUlP7k0NzWvRN/IwMBMPykzKSczP70osSCjEmoYVYyys9GHe8aOlwsAAAD//wMAUEsDBBQA BgAIAAAAIQDfEuNsuwEAADEDAAARAAgBZG9jUHJvcHMvY29yZS54bWwgogQBKKAAAQAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAACMkkFvGyEQhe+V+h8Qh94wsG7chq43UutGrRRLUbNVq94ojB3U hUVA7Pjfh8XeraPm0MtK7Ht8M/OG+urRdmgHIZreLTGfMYzAqV4bt13i7+01eY9RTNJp2fUOlvgA EV81r1/VygvVB7gNvYeQDESUSS4K5Zf4PiUvKI3qHqyMs+xwWdz0wcqUj2FLvVR/5BZoxdiCWkhS yyTpACR+IuITUqsJ6R9CVwBaUejAgkuR8hmnf70Jgo0vXijKmdOadPB5plO752ytjuLkfoxmMu73 +9l+XtrI/XP6c31zV0Ylxg1ZKcBNrZVQAWTqQ/O5vfta07Mfg6ghqmB8yrk3FeOcLSqOCGrB+k4m QA8+f7XgBN28kdZ/+IasDNtcAEVIKPWoUhZV5AtIDQEl+bsD1MEmIeN0TmV0sVL4vNiwuU7GtM5L 3hjQHw/NSu6MRmu1OkTpavqvYbgTYGeGV9IsqmKZziPwNhiXWx6mYYRXhC1afiEuuGDs1wQdTTmC so5jSKBRDlgc1zEqP+afVu01HnjzDCP8XcuZ4JdH3ugqIeeqE9Ce5vpP4ly8Zc+JI6ApTT9/5M0T AAAA//8DAFBLAwQUAAYACAAAACEA5NS1OvsCAAA2DgAAEgAAAHdvcmQvZm9udFRhYmxlLnhtbNxW TW/TMBi+I/Efoty3OFnXddW6aeso4sCEWBFnN3Ubi9iObHddr9sRiQsXDvAT+JBA4jJ+TWHX/QVe 22lLV3cUxAE1UtXkSd4n9pPn/dg7OGd5cEakooI3wngThQHhqehS3m+Ez9qtjVoYKI15F+eCk0Y4 Iio82L9/b29Y7wmuVQDxXNVZ2ggzrYt6FKk0IwyrTVEQDjd7QjKs4VL2I4bli0GxkQpWYE07NKd6 FCUIVcOSRq7CIno9mpJjkQ4Y4drGR5LkwCi4ymihJmzDVdiGQnYLKVKiFOyZ5Y6PYcqnNHFlgYjR VAolenoTNhO5FUWGCsJjZM9YHgYsrT/qcyFxJwfthnEl3C+FC4Z1jhmApyPWEbnFC8yFIjHcOsN5 I0SeI4wMQZphqYiePpg4uIcZzUcTFA+0cHhBdZpN4DMsqVmOu6VoH24MVAfN3hc6JAY7lCsokWTh ma15JLU8tV+iAAGeyUbMOyNnnAUd2pQRFZyQYfBUMMy9giQoQVW0hbZRBX4JnFXcNm4Lghz8bwTZ mgky3YiTaEVByqjY8KwuSFMMJCXSSLJEjB0QYBdkSKwYlXUW4znkpqlJyivF9uS7zP79vliHRDmE /PXXiwQdQXpUrCOMK5J1dkQT57QjqdcPCWrZlLASQJKstRA3V+9vrj4H129eXb9955VjlhXTM296 1LxlU9pybPsOF7otB6Q9Koi9nmssPXpOuo7X01Vs8YM+Y7pBDe2YhdzuKvEE8XYVCIBJwUWtXkTb OIPle1WZZItrJyZr1riZPD4NnjwUOqOp1QLn+gSmD+hHdtIoLXRz9Wl88WV88XV8eTm++PgHHlFD qpR7fs4Ty4eNB+ZbJq3WrLdWD5s7rePW0YItkhK5wxa7lmd1W/z48PL7t9deW5h13Tr8QlQdvGzG uDNXlutSvno+XaxYC7r8Ll0qhstEra5LaYQ5F5hJ7W8nU39FYaJLJHfy3SnTf1lSmphB31lWU8xg 6iqKGVTXq6aUo7va/wkAAP//AwBQSwMEFAAGAAgAAAAhAIhJ0TY8AgAA6A8AABQAAAB3b3JkL3dl YlNldHRpbmdzLnhtbOxX247aMBB9r9R/iPy+5L65aMOqdIVUqepD2X6ASQxYtT2RbUjZr++QkF3Y XVSQ0PYl4iHO2DPM+OQcj+/u/0jhbJg2HFRB/JFHHKZKqLhaFuTX4/QmJY6xVFVUgGIF2TJD7sef P901ecPmM2YtrjQORlEml2VBVtbWueuacsUkNSOomcLJBWhJLb7qpSup/r2ub0qQNbV8zgW3Wzfw vFuyD6PPiQKLBS/ZA5RryZRt/V3NBEYEZVa8Nn205pxoDeiq1lAyY7AeKbp4knL1HMaP3gSSvNRg YGFHWIzbZeTuQqG777UjKYgjy/zbUoGmc4E72PgRGeP2VXxj9k+nyXlVkDBLgzSOE6+dn0O1feAb nNtQgdAQd7caN+87W9je6j1bf/Ll6h3zI9Rv107AWpCv7JjPpNK7/7AvPgpBJ7jQPBUEPw0c1LTE ItpxCQIQK7q20KUhDjK7zHN+lNFlvvqw8ktc3RaEtuhueAxH5IdZGKVpPMBxyUdwDTg6dnxdcVEd Y5KGXpj6WdhC8ooLQRINbMivsf09G7qn2ePQW1twUAp2khUnfhDGWRIMHPlojrwHx23i+WkcpN0J M5wgZ55b1+DMSckKvCSKvSgL0kG0TjUO1wCg58O/RStF1YrTKOwAGVjygSzpQTo8Q1CxbjPstLyh 0bqo274GaU6qlh+FcZxF+BtU6/+p1l7Ldj0w1JZL/sSmoCcaGsN0ex+s4AfYGd2wL2aGN1fBplww nEHPg/v5+C8AAAD//wMAUEsDBBQABgAIAAAAIQA7uEXMIRkAAC6/AAAaAAAAd29yZC9zdHlsZXNX aXRoRWZmZWN0cy54bWzsXVtz27iSft+q/Q8qvexubWWiuyzX0ZySHDvJVMbjiZWZU3mjJTriRBJ1 SCpO5tdvo3EhiAsJkLQnPmdnHmIBJPqCrxsNoAH+7e9f97vOlzBJo/gw7/Z/6HU74WEdb6LDp3n3 w+rqxVm3k2bBYRPs4kM4734L0+7ff/zP//jbw3mafduFaQcaOKTnD8f1vLvNsuP5y5fpehvug/SH fbRO4jS+z35Yx/uX8f19tA5fPsTJ5uWg1+/hX8ckXodpCtQugsOXIO2y5vaxW2v7YM0bHvR6Zy/3 QXQQbegcxcfwAPzex8k+yNIf4uQTvJF8Ph1fAIfHIIvuol2UfQP+ehPRzJd595QczplUL4RU5J1z YOD8y37HHwa27c9SDZzTf/gbiSaogUn6yqt4fdqHhwzZe5mEO2A4PqTb6JjrrW5roI8tZ6lUYEnY h2N/pNET6nHp9FdJ8AB9zwk/HLXmDMrY0Jf2O6oHAqgcRmqL/Z5Dj5AmBA8uLBRpck5k8D3UU02O pIcjGGATg3qdxKejkOoYNWvt7eGzaIv4AQ/OehM0dVm01KsBzVfcboNj2O3s1+dvPx3iJLjbAUeg 8Q5BZPdH8E2beP0qvA9OuywlP5ObhP1kv/Cfq/iQpZ2H8yBdR9EKfBa0so+gwTeLQxp1oSYM0myR RoFcecnKSP2WPChXijfXaSY1uIw2UfclIZr+Ca99CXbz7mDASy4IE4WyXXD4xMvCw4vXS5mZeReK PtySojtod94Nkhe3C9LYS5SU/ytJfCzID7+QlWOwBuODZoL7LAQ/BF6PNLqLSAcPpuAB6Y/3J6Lf 4JTFjAg2AMTkZuGnonRwT+CsbukoAbXh/bt4/Tnc3GZQMe8iLSj88PYmieIE3O68O5sRmlB4G+6j N9FmE5JBiZV9OGyjTfj7Njx8SMNNXv7rFbpz1uI6Ph0yyj4RcpduLr+uwyPxlND0ISCdfE1eAJ8N 3SHRQYZOUc4NLVCoYuE/Ock+7UMjlW0YkGG0g/w/BaEBkUgTwIvXYfMmRs2bGDdvAsHbTBfT5lxA 8NS0R9Ai6gkSHTbhVwq/YgPoiYyYpa8YgGR/JYvXOo3hrMQwyBsaico3NGhWvmFAYrkYBtyVv2BA WfkLBkyVv2BAUPkLBrzYX7iP4+wQZ2EnC79mOlbtLxLfFiY+bxBSfm+sA/TaKpEh9pKTVxPiJeF9 mMC8BoIG1R7tMr6LUi+dkOc7y9NuF/q/dn3a3/lpB6lpVtQrMTt8Q7Oiyje8rAhpeJmRpDXdKVQy R9Xd8ZeKvegvHHvRX0baxX4yrqJsh5gtDR/6DQMVFih2boIk+JQEx22HRMY+pnJ7usvcWMVg6FQ7 prrNkpjMFysiN4itiaHXJnO5P26DNEqrCTVU/XXcIQbgo+sVme50XifRppK7MR0MNDXYfd7NLliH 23gHzr2z4mOCx/sgzy2dVlQy1xAJ76JP26wDM0ESY1cSm1j6ya4J2j7vnFL7m1hEqWrcqQ8nFijb G/853ESnPVeNw/RjQt2nRzcrJJDFchVRR+tPAoeH6hnUhPrjmu078E8jPf/2SR+7dAENDGu278A/ jSNrto/4KO9fb0/zCtZdhe8rbXvqbbsX8S5O7k87bgOV7mHqbcGChJOHmHobsWjfyUlMvS244D47 i/UalmpccOrdF7kf9aDi3R2UChqbuyzenaK4vb6HRN4dpNAaeNBq5ms9CHk73ffhl4jsNBGD1KMz +5CGo4AITyvNeWjRAFB1mjf+eoK5aiWVgcXnuVJ5e4D10TTsuFEbWizPlRrDE2rSB0zNBj4PMDUb AT0INRsKPQhZ8GGHuRgT3Yk0Hxw9aHm7ZTGKIeycPXPD4dJDIm+3bBw3HeIvi/XasaCPmw5UvDtI HzcdqLQ2bjrQ8u4g27jpQMsyatj7SPapIhhwIOQ9bsqEhPN2INSO83Yg1I7zdiDUjvN2INTceVcT ac95O9Dy9g1G5+1ACB/xie4EITl+dyDk7RuMzhtbKZ2ANlk84mORAxXvDtKdtwMV795RHKrwcw60 8BEfJCi0hKtzoNWO83Yg1I7zdiDUjvN2INSO83Yg1I7zdiDU3HlXE2nPeTvQ8vYNwqfKkbcDIW/3 IAjJztuBED7i4xuMzhut/tGdtwMV7w7SnbcDFe/eURyqcN4OtLw7SKElnLcDLXzEBwyMFoLbR6h2 nLeDRO04bwdC7ThvB0LtOG8HQs2ddzWR9py3Ay1v3yB8quy8HQh5uwdBSHbeDoS8fYPReaMxPrrz dqDi3UG683ag4t07ikMVfs6BlncHKbSE83aghXhp7LwdCOEjdQn5SNSO83aQqB3n7UCoHeftQKi5 864m0p7zdqDl7RuET5WdtwMhb/cgCMnO24GQt28wOm+0kUd33g5UvDtId94OVLx7R3Gownk70PLu IIWWcHUOtNpx3g6EEJiNnbcDIXykBiG0Ip9uasd5O0jUjvN2INTceVcTac95O9Dy9g3Cp8rO24GQ t3sQhGTn7UDI2zeQ1FzIF3XOaO1bQOCaZ8CzGpwJDiyd5EqQCfheTvEvHZeGDQlyCT0oWuDhKuIy jj933HLBhxaAOJOK7nZRjFng3zBLRzoPN5yWnC5Y/XLReUNPvGnvIaSKmTdwXFA+H4jnEclJQeAz +3aEM3pHnoxOWoMTgeQsJzvzhwfP38IJQHaOj7xMDvbBg3iKkhXjvi2jin/DIfcNf6bXGy6n/cUZ lYidiIzhIPz9Ln64OR3WmXiSPkKOP5LU5vDVpbXmWq3Z/HFKs/ckU/ntISdNG0zV45eYIsAOX47E j/eFw5egHcjKXuyiTwdyEpzTuwvSkLxJWqbnMolO4NQr/lM45zrvLpKIHn7E06vS7zUcy6WVqC7p qCo7h5H+KR1V5Wcz6FFVflJ23v0jePHTDWMEOQCGsMcq+lj0KkNRHw/yyv2aH6hE/g6gCa4AuctB EZ95OW/rYhskVO12QPQuRhfLCX2KAeJzGB6vgQzSIz/egZJT/HVcblC7WXzkxMg1Buy4xp8AVTR6 0suAZrTKNUlLLR6kZY38RcC7C+EmBOBuQOHGjgH3zwT4RBUBl4pEOCsIku/Ce2KXfZpztwU0gBZY ASqqErAP5/EpIwTefdlxVSK43JGco28V7eFOiuvwofM+3geYaGiCufoYcpoDfshgoAMeeGoC6oEV 1Ow0OLHkzS/kqDK6Mo5fikoPxA9qIt4A6wNc/IHJkABqRAbDNP5tx3Tu3zjMCLBQ0Xp/4zlqrb+l DkH1NFT+0Kr8IeXLrHzWMx7KH9ZUvkFlA6vKuEaYq2LOPlfZAMe2hiqjtymYnPCoTGVMnx4qG9VU WcEJjfo4VcqdECmwIA5ZLEHcAAVsqL6xFXHjMvUx3Xqob9yG+qY9jIxz9ZECi/q4dqzo4/Bs4i0n VvUx/6waLCv2UNykpuJ0F4Y9ygEF/9aJeaZWgVno3VzgaWsCo7IbCnxmFZgZriownWEwWHr081lN sWX/ggPefZSkGQkDyWBoMQ7srIaamVk1w65WMGuGqc1DM7OamtEtgA846BPsFrCGWDxYs8sALDM7 diRZHPnAA8nqPM9ybhk7Jb/wgodSDDP50gB9rnAkDorsfGfk4G3JbBQP5pZOSTv4CEWNziBcfoMs VXEIc8C7HZ3awR90YvnAQjM6d958DWhTUH8R7nY/BwnRHZ2qWB4lcTyoCmr7PRxFlabu4iyL9/QJ NASlPsHzuNYGQK0yM/QnEcKu7wPeicBO91pwch2ThTrNVPixagsUXDVt562wOiFmrkZesBAZUSxW nrMWlyYMM1HZD40nxTBncFYYjR25XsPSRLy/JYsu0OcY4JO/ybrKEm4UUI1tyexHESKXzkuE6bAo wmjcuggGCf6XWoUiApEW+6dSAkAktTv44y2Z/OIcDRUDSme18FeFA7SDR58eEv3CnREm/NRU/dmY M+zDZhlaYI6iocXOMqupVLaM+H5/VgxN28eLrvrlwIIXPierFMGMFx4Zt4AXfUaLeDFPZ0lVPeXD lTsAQE9glyEGDjNqiLEzzWoq1V1AzGSEO45iMtM+YnTlL4cWxIC8TTwMn6sKPEEogvcawr/cC5Hx mzjxYwyruLgcR7uMPurZd4URTV8DQJCxSariTElVrf5iU3VPRstApvO9tPPMarwwNujN2hjFCsrW VwxQ2eblAlJVh/H+tMedUFtjgM730s4zq/FT9niMC+qPaNDvdKf0zjjyekVuwmjhjzxowG4DsIva KpPOh+1aFl1mKNe63NePJTcfSZjc7Zn79S+aFFBkipu8eq+46VMYYfr0JjoByLyHWjKqy02UxUl6 TS4jUIdLWvdfsNVBao1iMuntVsaWjHEJnw8jV1c9+I80CF3js3ZXArDLX3X2fzXzTC9mRXFsSxh2 eVDtD+fFLjPYFbnrhF62h4PlaMy3e+BZrgeczorxdNaDI+lUJ0XrE5ush/gmieP7dvX2D11v/6jU m6d+CpAm4wJoQEC63/qofPm7LtPvRpkumag2eQCfOHISy8A/gHFD6kAJLq90hw9FCD4lqvF2Gdfm TWN9gwfOliLBP9YceQydFEkM0sxS7/zhJYILspMIoFf9CNlTh2LkAfZgCxvnYZIvDdp6gU12Yd0n 2sQPF3BvYBIXd3Fhe/ffK62inc1o0tcQHgi3AiAXV4fzXa0629OF1dcS46B4QQComKFVnRwcirUY F28tCKPIs6OLIT+x5bB838o2r/Rc0atwVbWyG3JRUYpGmfFqOqQt5R1h1yNzJE5uJgI61aO/K5Ak HlWJaVVTINFWvicgEQU+uu0Kz07kJ5c45/mIqp7v+TXPiUiSNKHMbrfKloQdZtJAhUsREbt3HxdK RVINH+r6fAvRJ9DU5MZ7Sq0ik/Qjig3FpuQxXR75uDpJs7llOWvAGITKQRZZiJJjLGVhSkSUkrt3 VJOPRRLACAlV5XEV4BWwbZgnp0Y7wlmXFuevJm+BNl2GAcNrJONRLUbLUJVPRl8197HaRRbcf8lY e3WjxWVQZDJRGbQ2K7QHxq4AIcMN7Lv1Ef4qOFhlp98UGJzK9+G58x7ns8E6Pe6pYZqlZ9PwoCUN Vy76W+xMJAq7WJd42MGmhoyjJ9AwTcWzaXjYkoYrV7yfWsM0+a+e1/LEMM3cs2l41JKGK5e5n1zD jKMnwDBN7rNpeNyShisX5Z9Kw483zr7REwqhyDTOsrGJ6aT2UqQc9/VnZ8UdSSxA4noqFaYCwUBe XP/Jxyi6BV0dgXjaMs2GtCFt0hLSmMpt8YuIgB950q9q80m8JU2/tGl42pKGWbbov6WGaaanTcNn LWmY5Vl+Pxp+wqiVZozaNDxrScMsG/JfQ8OFmaBYwXhLPtCmn3ODUQO/22YamOQJoPM82rAmwccW +2SxhGU9V4qyzMJ6ZaGFSYnS+LIM7cLsnp4xaz3N8d0rbdoNRcinHXWSLi2bPN/Ndod0kI+e3Qu/ Qt41FTCD1a1mx0gv4lMSwWeC4KwdWc9ymSMqr1hniuZdD552VyfiNoOZ5M/Q73npQIAq/jE3kx2S V8uh4mtZOTO6eZE6xk1JMigTxQ3AsmXVy1wQbiznXE+Jkzlns2TFPeRvV65TMNuTWYdMN55SgrX1 ta6ni8m8s9melffKFQAT7/V21w1617OvZN7ZLMbKO5Ot0ucV9F4vhazA+zX96KEaSCDrrO7pTI8S tJgerSwxPZldLz3WM72SteTrK82XQZFJjzw9qpJf6RQ8DxvEHC2fRNFNaXgA1wkc7bBMED1T5Nqc KVIlSHNebvRMEShyc7SGBKRCru6Qnt6A5JU8+6jwwBSOeyAp2wP9PqRGlj9hSnEqEOnP+PBqozIY 9pijsD4xOavgdMjyqOzCDkdjtjxgowL6YjkztidGA3p4105lNIEzTqUaG81mFTodD+EL6aVtjKc8 u8fG6aQ/qtDpZMyXcq1tzOCmiVI+poNZBacAsQqdnvUmFZyejUYVOj074wv/NllmA7ElDI9givt3 E02zSzLIRRXqTRjPMJimeq635FUIpks8+Gqhu01ShmBVohF5XuszIBXTPZ9weFotLrSBlpSZhBNC V0pWlp/TfBhbLd4YeH5j4ZnJYueZbQjcgcjAW0vD/mrxk4HHn4w8tgOah/Nc63DKdUtIPSmQ9Pz7 1cKcgO8OJDlWJzGmnHHSesy5Wrw39Nl7Y5+5i5D3Ck3GlrqlBVPQw+SVOUzmKcF2O1CWhEDVmO7E RlQ95xePT7PLlsio0q5gBhM3WzgXTN7d4mU2YR0VL6Z5cCucvr6ZxWv41ChIDR+BFNfLca0NK9d9 JXWTexjodVwl6cc8kpCCDpHdv8P053l3E+eH52YTygPIyl/9i4IQcXESxOagHwYZ/JuEIc63c40n eFcpSj3vjgbFfUhSi53RPIxRc6mAaZd1QcNrjmuDDfeKZZzqU3+CUxZaK9EKwropgE2OAdv08N2G ebjr1nAhipMVoS/iEUWYF+9AEUxDDSxZltd6kRwfkWt4IH1ljwjEpjN6zzJJm2xV4LkZAH/TgzRy t+hrfEQK89oedAsTr5EUtguhGvSFnoFBpGDTU70vKhIwGklny8doIJ1+fxORjrlXXTomtjwAyyFt E+kGA/WSEFIA7qWBdPplTUQ6bPXhXJeODfG+UuhuUdwbKDuKwWSGl2oKE8MCSUDhGu9ooc/8RLY7 /SImIjWLUnSpmTp8pZZFa8t7lE3NV3rAvqJRgCJSMUvXBlVbvCgymCpSnBy9eolEHxeaRFCE42kl c/cJ3B58k4DDxuuPBj28iHs7705nNFdf3j4lTyXBcd6FpPZlCFcK451QXxaH9Zbc7noMPmEB3DpK C/ZBAleikvjtG17MwFxaScDK7+AUVy7hdKLkdlmc3tlv4vyLolevcNUw2YPwu/EmtchKdglDxcOO wSd0KfYMX49uvDFdBu+lDu9lU3iTjAoQYtsevGdnbEwrgfdzgCN1x2zCxFfgnwdGyw6f8T2eR0Wq nlXz8ZU/Uuu7WNh2YcHvMwfhM3F/fFPnUUH1Wnd/r58KVF8xUWreFatGMJDDhln1QP4cPN3zcGrt Z4S5ZsZ/fH37swF7UFgaW7YQ0OoLuR/ZQq5PQFvPj3LI51el/Cth/pk41icZrfWJ4Ec2EWyGMtP8 h/nR/IaT53ptSPaOzCeoB3geDvQOvVWerQWf62AlDZfQy+YsK0MoCGVIWFlm+Og/mWHTFvJhHjKL MUzKwWed0dQovuolFoUiPnnju3TichWxIeezYFSmhQ/68PGB6sDHwsi9zGxhwiCp4/LDZAD37qL6 S0Jjvvzg9mWb/197oJkPgKzHvaOm/RCosANUhuDfdAT/ZrbiGsDWAxSyTjaB/wkJbriO8ZRYuF0G u10cH1bk9g81wZjV4e0WRjHkRXj5ahCp0QvxQSl948v1enGATOGLZIPJdHLFZhUst0bF1CrYwmeN cNFmcUgjuL1CFOCiFf3FPDt3caZcL5FnV+HlXMNkVTVlOm96QEyiRbvPuQsa6NtlAfFJO0NA/Q18 Ly8hMNWAnteYhlz7dT92fRYV2FtMenDNMjbOAFu4TJJcJXl1xfqIo5GtY1PbrkBfwUMJgckBgRvx eUAVaXh8IK82SS7bty4rD41KvhQnyc2yUQpTfaQpfTBQ+X1Nfxc+DohF+dYXzw6CqzhBbyzZA/KN 2Q81AVXePZryacMaLuMD93cKdrf0U4NIwzj3wRgKqrm3FWFS0f+stuEesk720SFO3hD/Q/jhhmGs xG+vGWvWaSa1tow2EcVJHqIOWGZBYV2Jle3oVwdx+T08vPhwSzi5g0bm3W344i3mmIIwteBF0fF7 eKfZE63p/DfU/Q/lVoliy4FVd1yAJPHLBTOzx4GbSDSieKM/FxBYM0TCgEwUzFDZp0/hL+WhxhCb d3++7dy8hjTMaC2jSy1HYKmFZAiUXqc2JQJ7fiiggCc2DdLxRK2hAkCF0VH4p4t4Tz7O+V5ctqa6 qOAAe4cBXo2WiIeQWwVPdVy0svNc/BI2Eik6cRldzInnNmi6abLWQR9VOcbATNILcVOVJibHZkzp pF2X2MxDLfbQs9D9JQG0ypuKB1bfynVrEi028GopltwPVeKg6P31q9Ik72+sZEbqk96XQ48PYwVz 5cmyFXZpjhuYauCD0X+EcA5a7QUJfSl7xGSTmoLlnAit0gBRRt8FpS3Y8h2V4YKkyVaPiJ6IlkVR 1clU0WHP2IN+SWe5Tux6q0K1orPvEeRyn5AjOGBHGMk0gbxrz93tTmEWwSfhNfznNSbY24eioobH i+FCTN3R6xNg4B8gJr2gG9Kq2XeS+RenVqTkdRJtNK6wpoNVJrawujzsGhszyi8v4aThJQUTC6q0 8FsOvmFZEPgnkXgx+M5DZp+vsi3jBK7dpt9wsX1AGnMzCE/4qUEgrmT40HsxOHaK6UHV75YlF1W/ jVuhdUlHMHPYhG+avf5bvdehu6CfZPU/50/kWca54EhiS82U1qzcZEeyCckDmlxumCqzWYkhmrLP ni8nk+WCLUPbLG8gvkHBnmCRadFzFhYbRldn/eUrYqQ4sbxgH1tfk90mnooqPK1PXMvjv/TH/xMA AAD//wMAUEsDBBQABgAIAAAAIQAF7br4JgIAAOMEAAAQAAgBZG9jUHJvcHMvYXBwLnhtbCCiBAEo oAABAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAALRUy24TMRTdI/EPo9k3M5OUUCLHFUpBRaIQ KdN2bTx3EguPbdm3UcOy3fAh7PgIPiY/wnUmjwmgLkBkdV85Pj4+d9j5faOTJfigrBmnRS9PEzDS VsrMx+l1+fbkLE0CClMJbQ2M0xWE9Jw/f8am3jrwqCAkBGHCOF0gulGWBbmARoQetQ11ausbgZT6 eWbrWkm4sPKuAYNZP8+HGdwjmAqqE7cHTFvE0RL/FrSyMvILN+XKEWHOSmicFgj8Q6Sje5XFhmX7 KistCl2qBnhRDKmxT9lUzCFwqrUBu7W+CrxfvCxY1sZsshBeSCQRedEfnFGjU2GvndNKCiSB+ZWS 3gZbY3IlpDJowyKJICzrTjFSaAbyzitc8Zxl3ZS9V4b4FPmAZW1IDL2Ye+EWROtV5LlP2UwKDRNS gtdCB2DZocAuQcRXngpFvNkSR0uQaH0S1Bd652GafBIBon7jdCm8EgZJxzjWJptYu4Celwo1YVOv zTdhd6wbq1NOAtEsBU8Otljrhx/rh2/rx6/rx+//6Qiib7X9d/B4qVZDutixuhuFwsea3gb/IPag K/bm3q3UHTm34bHGTxWP2Pxy/sQ2TpgVf1PO3iXkNA8YHxXJuNtOtNnncO1KexG3Zuue42LH97cK FzMnZHTm6Ys+efOwAZ0em9GmQEV+3iEeCuySrOZ1PJb+a+ZQ7WZ+b8Sdumm/WXReL6ffZoN2NVqC /ceE/wQAAP//AwBQSwECLQAUAAYACAAAACEAGbBocuYBAAAFCQAAEwAAAAAAAAAAAAAAAAAAAAAA W0NvbnRlbnRfVHlwZXNdLnhtbFBLAQItABQABgAIAAAAIQDCYJrz9AAAAE4CAAALAAAAAAAAAAAA AAAAAB8EAABfcmVscy8ucmVsc1BLAQItABQABgAIAAAAIQDCQ6xz7wEAALsIAAAcAAAAAAAAAAAA AAAAAEQHAAB3b3JkL19yZWxzL2RvY3VtZW50LnhtbC5yZWxzUEsBAi0AFAAGAAgAAAAhACR4odJM TwAAYrYBABEAAAAAAAAAAAAAAAAAdQoAAHdvcmQvZG9jdW1lbnQueG1sUEsBAi0AFAAGAAgAAAAh AD4oCXHlBQAAtBAAABAAAAAAAAAAAAAAAAAA8FkAAHdvcmQvaGVhZGVyMS54bWxQSwECLQAUAAYA CAAAACEAjR1Io4ACAACiCAAAEAAAAAAAAAAAAAAAAAADYAAAd29yZC9mb290ZXIxLnhtbFBLAQIt ABQABgAIAAAAIQAhQKSBCgIAAIkGAAASAAAAAAAAAAAAAAAAALFiAAB3b3JkL2Zvb3Rub3Rlcy54 bWxQSwECLQAUAAYACAAAACEAgTjG2AsCAACDBgAAEQAAAAAAAAAAAAAAAADrZAAAd29yZC9lbmRu b3Rlcy54bWxQSwECLQAUAAYACAAAACEABUbVi7oAAAAiAQAAGwAAAAAAAAAAAAAAAAAlZwAAd29y ZC9fcmVscy9oZWFkZXIxLnhtbC5yZWxzUEsBAi0ACgAAAAAAAAAhANZB0Y6mMAAApjAAABYAAAAA AAAAAAAAAAAAGGgAAHdvcmQvbWVkaWEvaW1hZ2UzLmpwZWdQSwECLQAUAAYACAAAACEAlrWt4pYG AABQGwAAFQAAAAAAAAAAAAAAAADymAAAd29yZC90aGVtZS90aGVtZTEueG1sUEsBAi0ACgAAAAAA AAAhAPO2SrdE9gAARPYAABUAAAAAAAAAAAAAAAAAu58AAHdvcmQvbWVkaWEvaW1hZ2UyLnBuZ1BL AQItAAoAAAAAAAAAIQBwcs4HbucAAG7nAAAVAAAAAAAAAAAAAAAAADKWAQB3b3JkL21lZGlhL2lt YWdlMS5wbmdQSwECLQAUAAYACAAAACEAhp7CMgMPAAAnPwAAEQAAAAAAAAAAAAAAAADTfQIAd29y ZC9zZXR0aW5ncy54bWxQSwECLQAUAAYACAAAACEAT5dJ4HwYAADMuwAADwAAAAAAAAAAAAAAAAAF jQIAd29yZC9zdHlsZXMueG1sUEsBAi0AFAAGAAgAAAAhABlcQqfDEwAAQPAAABIAAAAAAAAAAAAA AAAArqUCAHdvcmQvbnVtYmVyaW5nLnhtbFBLAQItABQABgAIAAAAIQB0Pzl6wgAAACgBAAAeAAAA AAAAAAAAAAAAAKG5AgBjdXN0b21YbWwvX3JlbHMvaXRlbTEueG1sLnJlbHNQSwECLQAUAAYACAAA ACEAwH00yeAAAABVAQAAGAAAAAAAAAAAAAAAAACnuwIAY3VzdG9tWG1sL2l0ZW1Qcm9wczEueG1s UEsBAi0AFAAGAAgAAAAhAKnIXKqMAAAA2gAAABMAAAAAAAAAAAAAAAAA5bwCAGN1c3RvbVhtbC9p dGVtMS54bWxQSwECLQAUAAYACAAAACEA3xLjbLsBAAAxAwAAEQAAAAAAAAAAAAAAAADKvQIAZG9j UHJvcHMvY29yZS54bWxQSwECLQAUAAYACAAAACEA5NS1OvsCAAA2DgAAEgAAAAAAAAAAAAAAAAC8 wAIAd29yZC9mb250VGFibGUueG1sUEsBAi0AFAAGAAgAAAAhAIhJ0TY8AgAA6A8AABQAAAAAAAAA AAAAAAAA58MCAHdvcmQvd2ViU2V0dGluZ3MueG1sUEsBAi0AFAAGAAgAAAAhADu4RcwhGQAALr8A ABoAAAAAAAAAAAAAAAAAVcYCAHdvcmQvc3R5bGVzV2l0aEVmZmVjdHMueG1sUEsBAi0AFAAGAAgA AAAhAAXtuvgmAgAA4wQAABAAAAAAAAAAAAAAAAAArt8CAGRvY1Byb3BzL2FwcC54bWxQSwUGAAAA ABgAGAAqBgAACuMCAAAA --Boundary_(ID_r/Imt8r/JUrLo5oMdJhX8Q)-- From jguichar@cisco.com Mon Jun 24 04:59:57 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C446411E812E for ; Mon, 24 Jun 2013 04:59:57 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -9.999 X-Spam-Level: X-Spam-Status: No, score=-9.999 tagged_above=-999 required=5 tests=[AWL=-0.600, BAYES_00=-2.599, J_CHICKENPOX_14=0.6, J_CHICKENPOX_47=0.6, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 7AFiEZonJByE for ; Mon, 24 Jun 2013 04:59:53 -0700 (PDT) Received: from rcdn-iport-6.cisco.com (rcdn-iport-6.cisco.com [173.37.86.77]) by ietfa.amsl.com (Postfix) with ESMTP id DDF6D11E8129 for ; Mon, 24 Jun 2013 04:59:52 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=2651; q=dns/txt; s=iport; t=1372075193; x=1373284793; h=from:to:cc:subject:date:message-id:in-reply-to: content-id:content-transfer-encoding:mime-version; bh=P40muZeONcNvJmiR7NNThgH+F4Dnp0apxMCnM6BevnM=; b=U4QTYwZhCHVzNBAy8tAMTRWDRSk156SmYxJs1r4Zr1q+bWCx4Z7CKxAE 39OypIa6QT1dJj+fFh2sum2gL/do1lYmVammJgd+wHOAka8WP31LI8sbz x+R9kACSGWHaL6JjOPISM+Ce+PfijJxBghBL1U1xh6Ltg3vR2izbCUCs3 g=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AmQFAN8zyFGtJV2Z/2dsb2JhbABbgwkxSb4ugQh+FnSCJQEEAQEBNy0HCxIBCCIUNwslAQEEAQ0FCIgGDLkhBI4MgRIxB4MCYQOhcocVgxCCKA X-IronPort-AV: E=Sophos;i="4.87,928,1363132800"; d="scan'208";a="226626020" Received: from rcdn-core-2.cisco.com ([173.37.93.153]) by rcdn-iport-6.cisco.com with ESMTP; 24 Jun 2013 11:59:52 +0000 Received: from xhc-aln-x15.cisco.com (xhc-aln-x15.cisco.com [173.36.12.89]) by rcdn-core-2.cisco.com (8.14.5/8.14.5) with ESMTP id r5OBxqR0002583 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Mon, 24 Jun 2013 11:59:52 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.80]) by xhc-aln-x15.cisco.com ([173.36.12.89]) with mapi id 14.02.0318.004; Mon, 24 Jun 2013 06:59:51 -0500 From: "Jim Guichard (jguichar)" To: Robert Raszuk , "adrian@olddog.co.uk" Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewCz9u0AABiS0oA= Date: Mon, 24 Jun 2013 11:59:51 +0000 Message-ID: <68B171751455884590F8E38E96416F363F551AFD@xmb-rcd-x01.cisco.com> In-Reply-To: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.2.3.120616 x-originating-ip: [10.98.130.212] Content-Type: text/plain; charset="us-ascii" Content-ID: Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Cc: "nsc@ietf.org" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 24 Jun 2013 11:59:57 -0000 Hi Robert, On 6/23/13 4:16 PM, "Robert Raszuk" wrote: >Hi Adrian, > >Great question ! (Just catching up ...) > >For me the definition is extremely simple ... Networks =3D services + >transport. > >So I somehow disagree with the notion of >draft-quinn-nsc-problem-statement that services are just data center >thing or that those are just firewalls, loadbalancers or worse carrier >grade NATs. For me services is anything BGP carries - that includes: >L3VPN regular PE, ASBR, L2VPNs, EVPNs, virtual hub PE, Internet, etc >... Jim> the intent of draft-quinn-nsc-problem-statement is not to imply that services are just for the data center, or that service chaining is only relevant in that environment; to the contrary, services and chaining may span multiple networks including the DC and WAN. Further, as I mentioned in a previous post, services are what the consumer buys and those services could be constructed using one or more "functions" where those functions could be a firewall as you said, but could also be a multitude of other things, some of which you mention above. > >Those with some history of L3VPNs recall very well the struggle to >chain firewalls on customer sites to make their VPN sites divided into >DMZ and secured planes. > >And just to cut some comments short .. this is not boiling the ocean. >Very simply service marking can accomplish any service discovery and >chaining (ability for specific data patter to be forced to go via such >service node+context within that node). Jim> this is certainly a valid way of forming basic service chains and we have included such mechanisms as part of the potential work items should a WG be formed.=20 > >Best, >R. > >> On Thu, Jun 20, 2013 at 1:23 PM, Adrian Farrel >>wrote: >> Hi, >> >> I know this cuts into use cases, but I think it will turn out to be >>important >> for clarity and scope. >> >> There is a risk that "we all know what we mean", but that no-one clearly >> articulates what "network service chaining" is. (It is a little like >>NFV where >> the definition also seems to be hard to pin down.) >> >> I don't want to rat-hole too much on definitions, but it would help if >>we can >> draw some lines. When is "service chaining" not "network service >>chaining"? >> >> Thanks, >> Adrian >> >> _______________________________________________ >> nsc mailing list >> nsc@ietf.org >> https://www.ietf.org/mailman/listinfo/nsc >_______________________________________________ >nsc mailing list >nsc@ietf.org >https://www.ietf.org/mailman/listinfo/nsc From jguichar@cisco.com Mon Jun 24 07:19:23 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9E45621E8117 for ; Mon, 24 Jun 2013 07:19:23 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.599 X-Spam-Level: X-Spam-Status: No, score=-10.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id a7XerP2RXy5l for ; Mon, 24 Jun 2013 07:19:18 -0700 (PDT) Received: from rcdn-iport-4.cisco.com (rcdn-iport-4.cisco.com [173.37.86.75]) by ietfa.amsl.com (Postfix) with ESMTP id 898FC21E811D for ; Mon, 24 Jun 2013 07:19:12 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=4501; q=dns/txt; s=iport; t=1372083554; x=1373293154; h=from:to:subject:date:message-id:in-reply-to:content-id: content-transfer-encoding:mime-version; bh=JQzkkx96CF9abrVFsnUvNMts973230xI/DHGqLRisV8=; b=G/fZOJRO6G+KBYVUv5KT3NcVdtoExkjm5tcYjv2ExdlUKUlTWFFnDmKK gN72qZzyZRjJYnEPmeBRLRcoBYg6I4I1B1N2t3mZ80zkDnW8RmmbGOGDp UFhXhX7H9djKer/H7PlR/PKBVpVtfnwa0i7Jz2iDS4hpZVklOU8QfCSa2 o=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AmQFAABUyFGtJV2c/2dsb2JhbABbgwkxSb4vgQiBBBZ0giMBAQEEAQEBax0BCBgKHS4LFBEBAQQBEgiIBgy5XwSPHjiDAmEDiGmZCYcVgxCCKA X-IronPort-AV: E=Sophos;i="4.87,928,1363132800"; d="scan'208";a="226755279" Received: from rcdn-core-5.cisco.com ([173.37.93.156]) by rcdn-iport-4.cisco.com with ESMTP; 24 Jun 2013 14:19:10 +0000 Received: from xhc-rcd-x09.cisco.com (xhc-rcd-x09.cisco.com [173.37.183.83]) by rcdn-core-5.cisco.com (8.14.5/8.14.5) with ESMTP id r5OEJAAl022797 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Mon, 24 Jun 2013 14:19:10 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.80]) by xhc-rcd-x09.cisco.com ([173.37.183.83]) with mapi id 14.02.0318.004; Mon, 24 Jun 2013 09:19:09 -0500 From: "Jim Guichard (jguichar)" To: "mohamed.boucadair@orange.com" , "adrian@olddog.co.uk" , "nsc@ietf.org" Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewABC13ID//+4ggP//KpNAgAbrYYA= Date: Mon, 24 Jun 2013 14:19:08 +0000 Message-ID: <68B171751455884590F8E38E96416F363F5521C2@xmb-rcd-x01.cisco.com> In-Reply-To: <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.2.3.120616 x-originating-ip: [10.98.130.212] Content-Type: text/plain; charset="iso-8859-1" Content-ID: Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 24 Jun 2013 14:19:23 -0000 Hi Med, I think we will find there are various models for how service functions are applied to service chains; the two most high-level models are (i) proxy functions, and (ii) non-proxy functions. We called these two models out in draft-quinn-nsh-00. In the former case it is true that the service function is essentially a black box but with a well defined interface and further that interface enables the blackbox to perform enhance service delivery. In this case it will be the job of the proxy to interpret the service headers and map them to a corresponding identifier that the service function can understand (for example a VLAN). In the latter case the service function itself will have the intelligence to interpret the service headers and therefore as part of our work we will need to define these headers, as well as the format and data presented to the service function. On 6/21/13 3:45 AM, "mohamed.boucadair@orange.com" wrote: >Hi Jim, all, > >I would even say: these functions are seen as black-boxes and no >assumption is to be made on the outcome of such functions. > >We tried to capture this and other assumptions in draft-boucadair-* > > o NLFs are handled as black-boxes; the technical characterization of > each NLF is not required. > o Not all NLFs can be characterized with a standard definition in > terms of technical description, detailed specification, etc. > o There is no global nor standard list of NLFs enabled in a given > administrative domain. The set of NLFs varies on a per deployment > basis. > o There is no global nor standard NLF chaining logic. The ordered > set of NLFs that need to be activated to deliver a given > connectivity service is specific to each administrative entity. > o The chaining of NLFs and the criteria to invoke some of them are > local to each administrative entity that operates a connectivity > network (also called administrative domain). > o NLF chaining logic and related policies should not be exposed > outside a given administrative domain. > o Several NLF chaining logics can be simultaneously enforced within > an administrative domain to meet business requirements. > >Cheers, >Med > >>-----Message d'origine----- >>De : nsc-bounces@ietf.org [mailto:nsc-bounces@ietf.org] De la part de Jim >>Guichard (jguichar) >>Envoy=E9 : jeudi 20 juin 2013 19:23 >>=C0 : adrian@olddog.co.uk; nsc@ietf.org >>Objet : Re: [nsc] What is a service? Why is NSC not SC? >> >>Hi Adrian, >> >>On 6/20/13 10:26 AM, "Adrian Farrel" wrote: >> >>>Thanks to Ken and Med for replies. >>> >>>Choosing to respond on Med's thread because he gave... >>> >>>> o NLF (Network-Located Function): refers to a function which is >>>> enabled in the network operated by an administrative entity. >>>>NLF >>>> can be for example: firewall (e.g., [RFC6092]), DPI (Deep Packet >>>> Inspection), LI (Lawful Intercept) module, NAT44 [RFC3022], >>>>NAT64 >>>> [RFC6146], DS-Lite AFTR [RFC6333], NPTv6 [RFC6296], HOST_ID >>>> injection, HTTP Header Enrichment function, TCP optimizer, etc. >>> >>>Something I pick out of this is that all the example functions are >>>packet-oriented functions (or possibly flow-oriented) and not >>>application >>>layer >>>services. I think this is significant and not accidental :-) >> >>Jim> actually the intent was that both packet-oriented and application >>layer service functions are within scope. It is certainly possible to >>envisage service chains that incorporate both of these types of functions >>and we should not restrict our work to one or the other but rather let >>the >>use cases dictate. >> >> >>>Can we capture that idea in a definition? >>> >>>I note that it is hinted at in Paul and Jim's BoF proposal text and it >>>is >>>there >>>because I asked them why this proposal was aimed at the Routing Area not >>>the >>>Applications Area. To which the answer was "because we want to steer >>>packets or >>>flows to the place where the next service will be applied." >>> >>>Adrian >>> >>>_______________________________________________ >>>nsc mailing list >>>nsc@ietf.org >>>https://www.ietf.org/mailman/listinfo/nsc >> >>_______________________________________________ >>nsc mailing list >>nsc@ietf.org >>https://www.ietf.org/mailman/listinfo/nsc From jguichar@cisco.com Mon Jun 24 07:35:39 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id B079421E813F for ; Mon, 24 Jun 2013 07:35:39 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.479 X-Spam-Level: X-Spam-Status: No, score=-10.479 tagged_above=-999 required=5 tests=[AWL=0.120, BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id lxlIC7BBbZVq for ; Mon, 24 Jun 2013 07:35:34 -0700 (PDT) Received: from rcdn-iport-5.cisco.com (rcdn-iport-5.cisco.com [173.37.86.76]) by ietfa.amsl.com (Postfix) with ESMTP id 8C27A21E8110 for ; Mon, 24 Jun 2013 07:35:34 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=4590; q=dns/txt; s=iport; t=1372084534; x=1373294134; h=from:to:subject:date:message-id:in-reply-to:content-id: content-transfer-encoding:mime-version; bh=baGu17RX7fAeInUgWb8mNt3EvtFNKXrQG558ZBoBzOI=; b=Sg0kCJRyeQbLeta21wteaMD6HmKGXokVhCHiJXFCymDONrBDBSuAmDHW 8nIXCCR3OUBwB/r6NyYesKTrjloIbzgKZbsv9BfZ8i3lHKJEe2MnEeeja ccJJpeb86RfAnCo2Cw9uigKZ4dMBFCShmFpZjZt0EWM0xmX8AXtvSdgiW k=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AmQFAKdYyFGtJV2a/2dsb2JhbABbgwkxSb4vgQiBBBZ0giMBAQEEAQEBax0BCBgKHS4LFBEBAQQBEgiIBgy5ewSPHjiDAmEDiGmZCYcVgxCCKA X-IronPort-AV: E=Sophos;i="4.87,928,1363132800"; d="scan'208";a="226684341" Received: from rcdn-core-3.cisco.com ([173.37.93.154]) by rcdn-iport-5.cisco.com with ESMTP; 24 Jun 2013 14:35:34 +0000 Received: from xhc-aln-x14.cisco.com (xhc-aln-x14.cisco.com [173.36.12.88]) by rcdn-core-3.cisco.com (8.14.5/8.14.5) with ESMTP id r5OEZXpE018250 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Mon, 24 Jun 2013 14:35:33 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.80]) by xhc-aln-x14.cisco.com ([173.36.12.88]) with mapi id 14.02.0318.004; Mon, 24 Jun 2013 09:35:33 -0500 From: "Jim Guichard (jguichar)" To: "mohamed.boucadair@orange.com" , "adrian@olddog.co.uk" , "nsc@ietf.org" Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewABC13ID//+4ggP//KpNAgAbv9oA= Date: Mon, 24 Jun 2013 14:35:32 +0000 Message-ID: <68B171751455884590F8E38E96416F363F5521FB@xmb-rcd-x01.cisco.com> In-Reply-To: <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.2.3.120616 x-originating-ip: [10.98.130.212] Content-Type: text/plain; charset="iso-8859-1" Content-ID: Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 24 Jun 2013 14:35:39 -0000 Hi Med, One further thought. All of the points that you make are well taken and honestly reflect one of the reasons why we have submitted our NSH proposal. Trying to define the characteristics and operation of each and every service function (NLF in your terminology) is imho looking at the problem from the wrong direction. What we want to be able to do is take the intelligence inherent within the network and present that information to whatever application/function is running within a service function so that it may take advantage of that data in whichever way it chooses. This requires that said intelligence is presented in a standardized format so that the developer of the service function can be sure of what information it can obtain from the network and how it might process that to provide the function it is designed to perform. It also provides the ability for one service function to interoperate with other service functions so as to build service chains. On 6/21/13 3:45 AM, "mohamed.boucadair@orange.com" wrote: >Hi Jim, all, > >I would even say: these functions are seen as black-boxes and no >assumption is to be made on the outcome of such functions. > >We tried to capture this and other assumptions in draft-boucadair-* > > o NLFs are handled as black-boxes; the technical characterization of > each NLF is not required. > o Not all NLFs can be characterized with a standard definition in > terms of technical description, detailed specification, etc. > o There is no global nor standard list of NLFs enabled in a given > administrative domain. The set of NLFs varies on a per deployment > basis. > o There is no global nor standard NLF chaining logic. The ordered > set of NLFs that need to be activated to deliver a given > connectivity service is specific to each administrative entity. > o The chaining of NLFs and the criteria to invoke some of them are > local to each administrative entity that operates a connectivity > network (also called administrative domain). > o NLF chaining logic and related policies should not be exposed > outside a given administrative domain. > o Several NLF chaining logics can be simultaneously enforced within > an administrative domain to meet business requirements. > >Cheers, >Med > >>-----Message d'origine----- >>De : nsc-bounces@ietf.org [mailto:nsc-bounces@ietf.org] De la part de Jim >>Guichard (jguichar) >>Envoy=E9 : jeudi 20 juin 2013 19:23 >>=C0 : adrian@olddog.co.uk; nsc@ietf.org >>Objet : Re: [nsc] What is a service? Why is NSC not SC? >> >>Hi Adrian, >> >>On 6/20/13 10:26 AM, "Adrian Farrel" wrote: >> >>>Thanks to Ken and Med for replies. >>> >>>Choosing to respond on Med's thread because he gave... >>> >>>> o NLF (Network-Located Function): refers to a function which is >>>> enabled in the network operated by an administrative entity. >>>>NLF >>>> can be for example: firewall (e.g., [RFC6092]), DPI (Deep Packet >>>> Inspection), LI (Lawful Intercept) module, NAT44 [RFC3022], >>>>NAT64 >>>> [RFC6146], DS-Lite AFTR [RFC6333], NPTv6 [RFC6296], HOST_ID >>>> injection, HTTP Header Enrichment function, TCP optimizer, etc. >>> >>>Something I pick out of this is that all the example functions are >>>packet-oriented functions (or possibly flow-oriented) and not >>>application >>>layer >>>services. I think this is significant and not accidental :-) >> >>Jim> actually the intent was that both packet-oriented and application >>layer service functions are within scope. It is certainly possible to >>envisage service chains that incorporate both of these types of functions >>and we should not restrict our work to one or the other but rather let >>the >>use cases dictate. >> >> >>>Can we capture that idea in a definition? >>> >>>I note that it is hinted at in Paul and Jim's BoF proposal text and it >>>is >>>there >>>because I asked them why this proposal was aimed at the Routing Area not >>>the >>>Applications Area. To which the answer was "because we want to steer >>>packets or >>>flows to the place where the next service will be applied." >>> >>>Adrian >>> >>>_______________________________________________ >>>nsc mailing list >>>nsc@ietf.org >>>https://www.ietf.org/mailman/listinfo/nsc >> >>_______________________________________________ >>nsc mailing list >>nsc@ietf.org >>https://www.ietf.org/mailman/listinfo/nsc From Ron_Parker@affirmednetworks.com Mon Jun 24 08:19:29 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 5940B11E8146 for ; Mon, 24 Jun 2013 08:19:29 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.598 X-Spam-Level: X-Spam-Status: No, score=-2.598 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HTML_MESSAGE=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id uA3rkhVvyPEO for ; Mon, 24 Jun 2013 08:19:23 -0700 (PDT) Received: from hub021-ca-7.exch021.serverdata.net (hub021-ca-7.exch021.serverdata.net [64.78.56.72]) by ietfa.amsl.com (Postfix) with ESMTP id F29D511E813F for ; Mon, 24 Jun 2013 08:19:11 -0700 (PDT) Received: from MBX021-W3-CA-2.exch021.domain.local ([10.254.4.78]) by HUB021-CA-7.exch021.domain.local ([10.254.4.109]) with mapi id 14.03.0123.003; Mon, 24 Jun 2013 08:19:10 -0700 From: Ron Parker To: "nsc@ietf.org" Thread-Topic: Comments on draft-quinn-nsh-00 Thread-Index: Ac5w7gaFkR09Y/lqTveMLRqlAXe6pQ== Date: Mon, 24 Jun 2013 15:19:10 +0000 Message-ID: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [50.203.66.100] Content-Type: multipart/alternative; boundary="_000_CDF2F015F4429F458815ED2A6C2B6B0B1A668097MBX021W3CA2exch_" MIME-Version: 1.0 Subject: [nsc] Comments on draft-quinn-nsh-00 X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 24 Jun 2013 15:20:54 -0000 --_000_CDF2F015F4429F458815ED2A6C2B6B0B1A668097MBX021W3CA2exch_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable I am new to this group and want to thank the authors and participants for t= he good work you've done so far. I feel that this work is very complement= ary to the Network Function Virtualization Forum's nascent network function= chaining work. Draft-quinn-nsh-00 and draft-quinn-nsc-problem-statement-= 00 are, however, appropriately more general in that mobile data services ar= e not assumed to be the de facto business problem and co-location of networ= k functions are not assumed to be within the data center. My specific questions and comments on draft-quinn-nsh-00 follow... General - I agree with some previous comments on the thread that "service" = is an overloaded term. I prefer the NFV's terms "network function" and "n= etwork function chaining". Let the business folks determine what set of f= unctions constitute a "service" that consumers or businesses will pay for. 2.2 Problem Statement; 1. Topological Dependencies - This would be a good p= lace to clarify whether service chains are unidirectional or bidirectional.= If bidirectional, is symmetry assumed (i.e. if from client to server net= work functions A, B, C are invoked, then from server to client, is it alway= s C, B, A? Must a classifier be at both ends of the chain? Must it be th= e same classifier at each end? 2.2 Problem Statement - It may be useful to discuss how standard IP routing= and NSC co-exist. For example, a deployment based on this technique may = utilize IP routing to ensure packets arrive at the two ends of a service ch= ain, but the service overlay is used to steer traffic between the two ends = of the chain. 3.1 NSH Actions; 1. Insert / Remove Service Header - this section introduce= s the concept of control plane to indicate which node is "end of the chain"= . This could also be accomplished via the management plane. Perhaps a d= iscussion of control plane and management plane could be added to section 2= .2? From the balance of the draft, it seems that control plane is deferre= d for a future proposal? 3.1 NSH Actions; 3. Update a Service Header - this section introduces the c= oncept of a service index. From context, I assume that this indicates the= previous node was considered to be index N out of M total nodes? Is it 0= -based or 1-based? 3.3 NSH Usage; 2. Service Chaining - this section introduces OAM messages, = but none are defined elsewhere in this document. Does this lend itself to= a separate draft? If so, should this document acknowledge that it won't = attempt to define OAM procedures? Perhaps it could acknowledge the possi= bility that the fields following the Base Header would be different for OAM= messages? 3.3 NSH Usage; 3. Metadata Sharing -It is mentioned that semantics of the m= etadata require coordination via a control plane, but this seems somewhat c= ontrary to the previous statement that a fixed size and format header is us= ed for simplicity. 3.4 NSH Proxy Nodes - it is mentioned that the proxy may deliver the packet= to the service node via a local attachment circuit. I suggest that it sho= uld be specified that the service node is required to deliver the packet ba= ck to the proxy (i.e., "bump-in-the-wire" forwarding paradigm) via the same= or different access circuit. Also, this mechanism could also be extended= to non-NSH tunnels (i..e, IP-in-IP or GRE). 4. Header Format; C bit - the text states that when C bit is set, there are= one or more contexts present. But then states that the format is fixed t= o 4 32-bit contexts. 4. Header Format - the distinction of network/service and platform/shared i= s not clear to me. Perhaps some usage examples could be added to help cla= rify? 4. Header Format - for deployment topologies that encompass access networ= ks (e.g., most), it is very common to wish to convey a subscriber identity = (e.g., IMSI or MSISDN in a mobile network) distinct from the IP address use= d in the packet. This may facilitate application of local policy. It ca= n also facilitate application of local policy in network functions chained = after a NAT function. Perhaps an optional 128-bit generic subscriber iden= tity can be considered in the NSH? On the other hand, I can see where ke= eping the NSH size minimal is extremely beneficial. One thought here woul= d be a mechanism that could be employed by stateful network functions using= flag bits to acknowledge that certain metadata has been received and is no= longer necessary in subsequent packets of the fully qualified flow. With= this approach, the 128-bit generic subscriber identity could be conveyed i= n the first N packets to the next node until a packet is received back from= the next node with the corresponding acknowledgement bit set. This appro= ach could be extended to other fields, including the 4 context fields alrea= dy defined. --_000_CDF2F015F4429F458815ED2A6C2B6B0B1A668097MBX021W3CA2exch_ Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable

I am new to this group and want to thank the authors= and participants for the good work you’ve done so far.   I= feel that this work is very complementary to the Network Function Virtuali= zation Forum’s nascent network function chaining work.   Draft-quinn-nsh-00 and draft-quinn-nsc-problem-statement-00 are, however, = appropriately more general in that mobile data services are not assumed to = be the de facto business problem and co-location of network functions are n= ot assumed to be within the data center.

My specific questions and comments on draft-quinn-ns= h-00 follow…

General – I agree with some previous comments = on the thread that “service” is an overloaded term.  = I prefer the NFV’s terms “network function” and “n= etwork function chaining”.   Let the business folks determi= ne what set of functions constitute a “service” that consumers or businesses will pay for.

2.2 Problem Statement; 1. Topological Dependencies &= #8211; This would be a good place to clarify whether service chains are uni= directional or bidirectional.   If bidirectional, is symmetry ass= umed (i.e. if from client to server network functions A, B, C are invoked, then from server to client, is it always C, B, A?&nbs= p; Must a classifier be at both ends of the chain?   Must it be t= he same classifier at each end?

2.2 Problem Statement – It may be useful to di= scuss how standard IP routing and NSC co-exist.   For example, a = deployment based on this technique may utilize IP routing to ensure packets= arrive at the two ends of a service chain, but the service overlay is used to steer traffic between the two ends of the chain= .

3.1 NSH Actions; 1. Insert / Remove Service Header &= #8211; this section introduces the concept of control plane to indicate whi= ch node is “end of the chain”.   This could also be a= ccomplished via the management plane.   Perhaps a discussion of control plane and management plane could be added to section 2.2? &nb= sp; From the balance of the draft, it seems that control plane is deferred = for a future proposal?

3.1 NSH Actions; 3. Update a Service Header – = this section introduces the concept of a service index.   From co= ntext, I assume that this indicates the previous node was considered to be = index N out of M total nodes?   Is it 0-based or 1-based?

3.3 NSH Usage; 2. Service Chaining – this sect= ion introduces OAM messages, but none are defined elsewhere in this documen= t.   Does this lend itself to a separate draft?   If so= , should this document acknowledge that it won’t attempt to define OAM procedures?    Perhaps it could acknowledge the = possibility that the fields following the Base Header would be different fo= r OAM messages?

3.3 NSH Usage; 3. Metadata Sharing –It is ment= ioned that semantics of the metadata require coordination via a control pla= ne, but this seems somewhat contrary to the previous statement that a fixed= size and format header is used for simplicity.

3.4 NSH Proxy Nodes – it is mentioned that the= proxy may deliver the packet to the service node via a local attachment ci= rcuit.  I suggest that it should be specified that the service node is= required to deliver the packet back to the proxy (i.e., “bump-in-the-wire” forwarding paradigm) via the same or= different access circuit.   Also, this mechanism could also be e= xtended to non-NSH tunnels (i..e, IP-in-IP or GRE).

4. Header Format; C bit – the text states that= when C bit is set, there are one or more contexts present.   But= then states that the format is fixed to 4 32-bit contexts.

4. Header Format – the distinction of network/= service and platform/shared is not clear to me.   Perhaps some us= age examples could be added to help clarify?

4.  Header Format – for deployment topolo= gies that  encompass access networks (e.g., most), it is very common t= o wish to convey a subscriber identity (e.g., IMSI or MSISDN in a mobile ne= twork) distinct from the IP address used in the packet.   This may facilitate application of local policy.   It can also f= acilitate application of local policy in network functions chained after a = NAT function.   Perhaps an optional 128-bit generic subscriber id= entity can be considered in the NSH?    On the other hand, I can see where keeping the NSH size minimal is extremely beneficial= .   One thought here would be a mechanism that could be employed = by stateful network functions using flag bits to acknowledge that certain m= etadata has been received and is no longer necessary in subsequent packets of the fully qualified flow.   W= ith this approach, the 128-bit generic subscriber identity could be conveye= d in the first N packets to the next node until a packet is received back f= rom the next node with the corresponding acknowledgement bit set.   This approach could be extended to other fields, incl= uding the 4 context fields already defined.

 

--_000_CDF2F015F4429F458815ED2A6C2B6B0B1A668097MBX021W3CA2exch_-- From diego@tid.es Mon Jun 24 10:09:39 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 061D821E8106 for ; Mon, 24 Jun 2013 10:09:39 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.471 X-Spam-Level: X-Spam-Status: No, score=-6.471 tagged_above=-999 required=5 tests=[AWL=0.128, BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id XSu2gIBDXhpw for ; Mon, 24 Jun 2013 10:09:34 -0700 (PDT) Received: from tidos.tid.es (tidos.tid.es [195.235.93.44]) by ietfa.amsl.com (Postfix) with ESMTP id EE56111E8161 for ; Mon, 24 Jun 2013 10:09:33 -0700 (PDT) Received: from sbrightmailg01.hi.inet (sbrightmailg01.hi.inet [10.95.64.104]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTP id <0MOW00BBHQBPE1@tid.hi.inet> for nsc@ietf.org; Mon, 24 Jun 2013 19:09:32 +0200 (MEST) Received: from tid (tid.hi.inet [10.95.64.10]) by sbrightmailg01.hi.inet (Symantec Messaging Gateway) with SMTP id DD.3A.03142.C4D78C15; Mon, 24 Jun 2013 19:09:32 +0200 (CEST) Received: from correo.tid.es (mailhost.hi.inet [10.95.64.100]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTP id <0MOW00BCIQBWE1@tid.hi.inet> for nsc@ietf.org; Mon, 24 Jun 2013 19:09:32 +0200 (MEST) Received: from EX10-MB2-MAD.hi.inet ([169.254.2.38]) by EX10-HTCAS6-MAD.hi.inet ([::1]) with mapi id 14.02.0328.009; Mon, 24 Jun 2013 19:09:32 +0200 Date: Mon, 24 Jun 2013 17:08:30 +0000 From: "Diego R. Lopez" In-reply-to: <68B171751455884590F8E38E96416F363F5521FB@xmb-rcd-x01.cisco.com> X-Originating-IP: [10.95.64.115] To: "Jim Guichard (jguichar)" Message-id: Content-id: <5D2764A43C37C041B6C832C894D20D40@hi.inet> MIME-version: 1.0 Content-type: text/plain; charset=iso-8859-1 Content-language: en-US Content-transfer-encoding: quoted-printable Accept-Language: en-US, es-ES Thread-topic: [nsc] What is a service? Why is NSC not SC? Thread-index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewAAIKxYAABiYCAAAeHsEAAKUyZAAABWDeAA== X-AuditID: 0a5f4068-b7f128e000000c46-a9-51c87d4cca06 X-MS-Has-Attach: X-MS-TNEF-Correlator: X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFmpjkeLIzCtJLcpLzFFi42Lhinfg0vWpPRFo8OeppsXxffvZHBg9liz5 yRTAGMVlk5Kak1mWWqRvl8CVcf7/IZaCE8IVr95fZWxgfMnfxcjJISFgIvF/9zlmCFtM4sK9 9WxdjFwcQgIbGSX+tE1ihHC+MUocPbOGCcKZxigx+9R2NpAWFgFVieZDL8BsNiD7UfNvdhBb WMBCYtvNf0ANHBycAr4STxepQGxQkPhz7jELiC0iYCTRfek9K8hMZoF+RolXT76CncEr4C1x qX0imM0sYCbx8fxPFoi4oMSPyfdYIOI6Er3fv0HViEs0t96EimtLPHl3gRXEZhSQlXg3fz4r yA0iApYSDZ/kIPZGSGzsf84GcY+AxJI956G+F5V4+fgfWKuQgI/E+z+3WCcwSsxCcsUsJFfM QnLFLCRXzEJyxQJG1lWMYsVJRZnpGSW5iZk56QaGehmZepl5qSWbGCFxl7GDcflOlUOMAhyM Sjy8M2JOBAqxJpYVV+YeYpTgYFYS4Q0VAQrxpiRWVqUW5ccXleakFh9iZOLglGpgnCH5XTx3 3cOnjL7299oiVi6+vdUppO3ojOsKPseW3BPa78Ly7cRzx9wVJincSeev/xST0Pu+efqGgGmP OmyOO1V9/p7uYHDc7Ow0h3k3f8SW5D2+82rJ1e1SreKvZ32+Jr9kOeviIx/Z186ojrhyn/tp 7T2L5FfamvvmRNl/2jHLtS/i4rHCtXlKLMUZiYZazEXFiQCFYCr+mQIAAA== References: <68B171751455884590F8E38E96416F363F5521FB@xmb-rcd-x01.cisco.com> Cc: "nsc@ietf.org" , "adrian@olddog.co.uk" , "mohamed.boucadair@orange.com" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 24 Jun 2013 17:09:39 -0000 Hi Jim, On 24 Jun 2013, at 16:35 , Jim Guichard (jguichar) wrote: > > One further thought. All of the points that you make are well taken and > honestly reflect one of the reasons why we have submitted our NSH > proposal. Trying to define the characteristics and operation of each and > every service function (NLF in your terminology) is imho looking at the > problem from the wrong direction. What we want to be able to do is take > the intelligence inherent within the network and present that information > to whatever application/function is running within a service function so > that it may take advantage of that data in whichever way it chooses. This > requires that said intelligence is presented in a standardized format so > that the developer of the service function can be sure of what informatio= n > it can obtain from the network and how it might process that to provide > the function it is designed to perform. So this requires a standardized network-to-SF interface. I assume you refer= here to a mechanism by which the SF could learn what data it can retrieve = from the network once deployed in a certain location, as well as defining t= he mechanisms for attaching a service to that location (whether by means of= an address/port pair, a URL, a tunnel endpoint...) > It also provides the ability for one service function to interoperate wit= h > other service functions so as to build service chains. And this requires a standardized SF metadata interface, so each SF can anno= unce what it expects as input and provides as output, so an orchestrator ca= n statically or even dynamically build the appropriate chain. And this open= the question on whether chaining is linear and unidirectional, so we could= speak of pipelining, or it is foreseeable more complex composition structu= res. Be goode, -- "Esta vez no fallaremos, Doctor Infierno" Dr Diego R. Lopez Telefonica I+D http://people.tid.es/diego.lopez/ e-mail: diego@tid.es Tel: +34 913 129 041 Mobile: +34 682 051 091 ----------------------------------------- ________________________________ Este mensaje se dirige exclusivamente a su destinatario. Puede consultar nu= estra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el enl= ace situado m=E1s abajo. This message is intended exclusively for its addressee. We only send and re= ceive email on the basis of the terms set out at: http://www.tid.es/ES/PAGINAS/disclaimer.aspx From mohamed.boucadair@orange.com Mon Jun 24 23:19:35 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id A84E421F8D0D for ; Mon, 24 Jun 2013 23:19:35 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.133 X-Spam-Level: X-Spam-Status: No, score=-2.133 tagged_above=-999 required=5 tests=[AWL=0.115, BAYES_00=-2.599, HELO_EQ_FR=0.35, UNPARSEABLE_RELAY=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 81MQ1J0CtrDq for ; Mon, 24 Jun 2013 23:19:31 -0700 (PDT) Received: from relais-inet.francetelecom.com (relais-ias92.francetelecom.com [193.251.215.92]) by ietfa.amsl.com (Postfix) with ESMTP id 49D5F21F9F8C for ; Mon, 24 Jun 2013 23:19:31 -0700 (PDT) Received: from omfedm06.si.francetelecom.fr (unknown [xx.xx.xx.2]) by omfedm09.si.francetelecom.fr (ESMTP service) with ESMTP id B55602DC9CB; Tue, 25 Jun 2013 08:19:18 +0200 (CEST) Received: from PUEXCH81.nanterre.francetelecom.fr (unknown [10.101.44.34]) by omfedm06.si.francetelecom.fr (ESMTP service) with ESMTP id 929EE27C058; Tue, 25 Jun 2013 08:19:18 +0200 (CEST) Received: from PUEXCB1B.nanterre.francetelecom.fr ([10.101.44.13]) by PUEXCH81.nanterre.francetelecom.fr ([10.101.44.34]) with mapi; Tue, 25 Jun 2013 08:19:18 +0200 From: To: "Jim Guichard (jguichar)" , "adrian@olddog.co.uk" , "nsc@ietf.org" Date: Tue, 25 Jun 2013 08:19:16 +0200 Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewABC13ID//+4ggP//KpNAgAbv9oD//wsUIA== Message-ID: <94C682931C08B048B7A8645303FDC9F36EDB5C7B2A@PUEXCB1B.nanterre.francetelecom.fr> References: <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> <68B171751455884590F8E38E96416F363F5521FB@xmb-rcd-x01.cisco.com> In-Reply-To: <68B171751455884590F8E38E96416F363F5521FB@xmb-rcd-x01.cisco.com> Accept-Language: fr-FR Content-Language: fr-FR X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: fr-FR Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-PMX-Version: 5.6.1.2065439, Antispam-Engine: 2.7.2.376379, Antispam-Data: 2013.6.25.45432 Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 25 Jun 2013 06:19:35 -0000 Hi Jim, I think we are sync on the required protocol extensions and the merits of t= he approach. On important aspect to consider is provisioning and policy enforcement: as = you can see in draft-boucadair-*, we separated these aspects from the actua= l handling of packets and the behavior of involved nodes. I would vote to i= nclude the work on the Information Model (data model may not be in scope in= a first step since DMs will depend on the favorite provisioning protocol a= network is using).=20 Cheers, Med >-----Message d'origine----- >De=A0: Jim Guichard (jguichar) [mailto:jguichar@cisco.com] >Envoy=E9=A0: lundi 24 juin 2013 16:36 >=C0=A0: BOUCADAIR Mohamed OLNC/OLN; adrian@olddog.co.uk; nsc@ietf.org >Objet=A0: Re: [nsc] What is a service? Why is NSC not SC? > >Hi Med, > >One further thought. All of the points that you make are well taken and >honestly reflect one of the reasons why we have submitted our NSH >proposal. Trying to define the characteristics and operation of each and >every service function (NLF in your terminology) is imho looking at the >problem from the wrong direction. What we want to be able to do is take >the intelligence inherent within the network and present that information >to whatever application/function is running within a service function so >that it may take advantage of that data in whichever way it chooses. This >requires that said intelligence is presented in a standardized format so >that the developer of the service function can be sure of what information >it can obtain from the network and how it might process that to provide >the function it is designed to perform. It also provides the ability for >one service function to interoperate with other service functions so as to >build service chains. > >On 6/21/13 3:45 AM, "mohamed.boucadair@orange.com" > wrote: > >>Hi Jim, all, >> >>I would even say: these functions are seen as black-boxes and no >>assumption is to be made on the outcome of such functions. >> >>We tried to capture this and other assumptions in draft-boucadair-* >> >> o NLFs are handled as black-boxes; the technical characterization of >> each NLF is not required. >> o Not all NLFs can be characterized with a standard definition in >> terms of technical description, detailed specification, etc. >> o There is no global nor standard list of NLFs enabled in a given >> administrative domain. The set of NLFs varies on a per deployment >> basis. >> o There is no global nor standard NLF chaining logic. The ordered >> set of NLFs that need to be activated to deliver a given >> connectivity service is specific to each administrative entity. >> o The chaining of NLFs and the criteria to invoke some of them are >> local to each administrative entity that operates a connectivity >> network (also called administrative domain). >> o NLF chaining logic and related policies should not be exposed >> outside a given administrative domain. >> o Several NLF chaining logics can be simultaneously enforced within >> an administrative domain to meet business requirements. >> >>Cheers, >>Med >> >>>-----Message d'origine----- >>>De : nsc-bounces@ietf.org [mailto:nsc-bounces@ietf.org] De la part de Ji= m >>>Guichard (jguichar) >>>Envoy=E9 : jeudi 20 juin 2013 19:23 >>>=C0 : adrian@olddog.co.uk; nsc@ietf.org >>>Objet : Re: [nsc] What is a service? Why is NSC not SC? >>> >>>Hi Adrian, >>> >>>On 6/20/13 10:26 AM, "Adrian Farrel" wrote: >>> >>>>Thanks to Ken and Med for replies. >>>> >>>>Choosing to respond on Med's thread because he gave... >>>> >>>>> o NLF (Network-Located Function): refers to a function which is >>>>> enabled in the network operated by an administrative entity. >>>>>NLF >>>>> can be for example: firewall (e.g., [RFC6092]), DPI (Deep Packe= t >>>>> Inspection), LI (Lawful Intercept) module, NAT44 [RFC3022], >>>>>NAT64 >>>>> [RFC6146], DS-Lite AFTR [RFC6333], NPTv6 [RFC6296], HOST_ID >>>>> injection, HTTP Header Enrichment function, TCP optimizer, etc. >>>> >>>>Something I pick out of this is that all the example functions are >>>>packet-oriented functions (or possibly flow-oriented) and not >>>>application >>>>layer >>>>services. I think this is significant and not accidental :-) >>> >>>Jim> actually the intent was that both packet-oriented and application >>>layer service functions are within scope. It is certainly possible to >>>envisage service chains that incorporate both of these types of function= s >>>and we should not restrict our work to one or the other but rather let >>>the >>>use cases dictate. >>> >>> >>>>Can we capture that idea in a definition? >>>> >>>>I note that it is hinted at in Paul and Jim's BoF proposal text and it >>>>is >>>>there >>>>because I asked them why this proposal was aimed at the Routing Area no= t >>>>the >>>>Applications Area. To which the answer was "because we want to steer >>>>packets or >>>>flows to the place where the next service will be applied." >>>> >>>>Adrian >>>> >>>>_______________________________________________ >>>>nsc mailing list >>>>nsc@ietf.org >>>>https://www.ietf.org/mailman/listinfo/nsc >>> >>>_______________________________________________ >>>nsc mailing list >>>nsc@ietf.org >>>https://www.ietf.org/mailman/listinfo/nsc From jguichar@cisco.com Tue Jun 25 04:04:20 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6F06E21F9DC1 for ; Tue, 25 Jun 2013 04:04:20 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.499 X-Spam-Level: X-Spam-Status: No, score=-10.499 tagged_above=-999 required=5 tests=[AWL=0.100, BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 1oTBHud4nft0 for ; Tue, 25 Jun 2013 04:04:15 -0700 (PDT) Received: from rcdn-iport-8.cisco.com (rcdn-iport-8.cisco.com [173.37.86.79]) by ietfa.amsl.com (Postfix) with ESMTP id 160A421F9D38 for ; Tue, 25 Jun 2013 04:04:13 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=2967; q=dns/txt; s=iport; t=1372158253; x=1373367853; h=from:to:cc:subject:date:message-id:in-reply-to: content-id:content-transfer-encoding:mime-version; bh=jVE+JBfrcXWeCguX/YZDzvh0ucjY1e8JVsAyTF2rmY4=; b=It2C8Y0kKw+waL7fTCla4TMYexfYnVQj0C6q9QZMIkcaKvwOAeHN1KBR fUCbV79k6rh6z89lD4LqyJNhfCoCZm+vPqIxLpnseQstGnMkFD4zlHXgz ZagpcsDt336xT1LVGcEBZYtSx8xCPm23rl5istSpmKitpeiaYagCtIoFz s=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: Aj8FADx4yVGtJXG+/2dsb2JhbABagwkxSb4qgRCBARZ0giMBAQEDAXkSAQgYClYlAgQOBQiIAAYMunuOAxB/AiwFB4MCYQOIaZADiQaHFYMQgXE3 X-IronPort-AV: E=Sophos;i="4.87,936,1363132800"; d="scan'208";a="227101011" Received: from rcdn-core2-3.cisco.com ([173.37.113.190]) by rcdn-iport-8.cisco.com with ESMTP; 25 Jun 2013 11:04:12 +0000 Received: from xhc-aln-x07.cisco.com (xhc-aln-x07.cisco.com [173.36.12.81]) by rcdn-core2-3.cisco.com (8.14.5/8.14.5) with ESMTP id r5PB4CHZ001725 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Tue, 25 Jun 2013 11:04:12 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.80]) by xhc-aln-x07.cisco.com ([173.36.12.81]) with mapi id 14.02.0318.004; Tue, 25 Jun 2013 06:04:12 -0500 From: "Jim Guichard (jguichar)" To: "Diego R. Lopez" Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewABC13ID//+4ggP//KpNAgAbv9oCAAG3NAIAA6X4A Date: Tue, 25 Jun 2013 11:04:11 +0000 Message-ID: <68B171751455884590F8E38E96416F363F552B23@xmb-rcd-x01.cisco.com> In-Reply-To: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.2.3.120616 x-originating-ip: [10.98.130.212] Content-Type: text/plain; charset="iso-8859-1" Content-ID: <8C0103403F378F40BD0F2B67778CA075@emea.cisco.com> Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Cc: "nsc@ietf.org" , "adrian@olddog.co.uk" , "mohamed.boucadair@orange.com" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 25 Jun 2013 11:04:20 -0000 Hi Diego, On 6/24/13 1:08 PM, "Diego R. Lopez" wrote: >Hi Jim, > >On 24 Jun 2013, at 16:35 , Jim Guichard (jguichar) wrote: >> >> One further thought. All of the points that you make are well taken and >> honestly reflect one of the reasons why we have submitted our NSH >> proposal. Trying to define the characteristics and operation of each and >> every service function (NLF in your terminology) is imho looking at the >> problem from the wrong direction. What we want to be able to do is take >> the intelligence inherent within the network and present that >>information >> to whatever application/function is running within a service function so >> that it may take advantage of that data in whichever way it chooses. >>This >> requires that said intelligence is presented in a standardized format so >> that the developer of the service function can be sure of what >>information >> it can obtain from the network and how it might process that to provide >> the function it is designed to perform. > >So this requires a standardized network-to-SF interface. I assume you >refer here to a mechanism by which the SF could learn what data it can >retrieve from the network once deployed in a certain location, as well as >defining the mechanisms for attaching a service to that location (whether >by means of an address/port pair, a URL, a tunnel endpoint...) Jim> yes. > >> It also provides the ability for one service function to interoperate >>with >> other service functions so as to build service chains. > >And this requires a standardized SF metadata interface, so each SF can >announce what it expects as input and provides as output, so an >orchestrator can statically or even dynamically build the appropriate >chain. And this open the question on whether chaining is linear and >unidirectional, so we could speak of pipelining, or it is foreseeable >more complex composition structures. Jim> yes although the SF may or may not announce its capabilities (in the latter case the orchestrator would be told directly). In terms of service chains we have to work through both unidirectional and bidirectional as well as complex chains (where traffic may be diverted mid-chain for example on to a different chain). > >Be goode, > > >-- >"Esta vez no fallaremos, Doctor Infierno" > >Dr Diego R. Lopez >Telefonica I+D >http://people.tid.es/diego.lopez/ > >e-mail: diego@tid.es >Tel: +34 913 129 041 >Mobile: +34 682 051 091 >----------------------------------------- > > >________________________________ > >Este mensaje se dirige exclusivamente a su destinatario. Puede consultar >nuestra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el = enlace >situado m=E1s abajo. >This message is intended exclusively for its addressee. We only send and >receive email on the basis of the terms set out at: >http://www.tid.es/ES/PAGINAS/disclaimer.aspx From jguichar@cisco.com Tue Jun 25 04:27:35 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 4E9DA21F9EE6 for ; Tue, 25 Jun 2013 04:27:35 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.599 X-Spam-Level: X-Spam-Status: No, score=-10.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id mJLblLz5qBFl for ; Tue, 25 Jun 2013 04:27:30 -0700 (PDT) Received: from rcdn-iport-3.cisco.com (rcdn-iport-3.cisco.com [173.37.86.74]) by ietfa.amsl.com (Postfix) with ESMTP id 2B98921F9A2A for ; Tue, 25 Jun 2013 04:27:26 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=5859; q=dns/txt; s=iport; t=1372159646; x=1373369246; h=from:to:subject:date:message-id:in-reply-to:content-id: content-transfer-encoding:mime-version; bh=QB1V9xgWBI1dNSGu8p5genw0B/PxpZf2zhuXSiN44U0=; b=juGIQgWTauhnuziChqgnmA+/AwmcZyYavuvVEJ3clDffLpyJc4p2XGPS aIGJwoyfPotBfHNvKkQQoHxW3yqynVat3UyhkerhAzdzHKpLfuKLfYDRe zivSgH4+sVmgvZdYkMLzZXgFedvZ7bJtPP/Z3jNnED9124uSJiP6MVGXO o=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AkAFAHR9yVGtJV2Z/2dsb2JhbABagwkxSb4qgRCBARZ0giMBAQEEAQEBax0BCBgKHS4LFBECBAESCIgGDLp0BI8UOIMCYQOIaZkJhxWDEIIo X-IronPort-AV: E=Sophos;i="4.87,936,1363132800"; d="scan'208";a="227134522" Received: from rcdn-core-2.cisco.com ([173.37.93.153]) by rcdn-iport-3.cisco.com with ESMTP; 25 Jun 2013 11:27:23 +0000 Received: from xhc-rcd-x07.cisco.com (xhc-rcd-x07.cisco.com [173.37.183.81]) by rcdn-core-2.cisco.com (8.14.5/8.14.5) with ESMTP id r5PBRNdA028477 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Tue, 25 Jun 2013 11:27:23 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.80]) by xhc-rcd-x07.cisco.com ([173.37.183.81]) with mapi id 14.02.0318.004; Tue, 25 Jun 2013 06:27:23 -0500 From: "Jim Guichard (jguichar)" To: "mohamed.boucadair@orange.com" , "adrian@olddog.co.uk" , "nsc@ietf.org" Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewABC13ID//+4ggP//KpNAgAbv9oD//wsUIIACUq+A Date: Tue, 25 Jun 2013 11:27:22 +0000 Message-ID: <68B171751455884590F8E38E96416F363F553681@xmb-rcd-x01.cisco.com> In-Reply-To: <94C682931C08B048B7A8645303FDC9F36EDB5C7B2A@PUEXCB1B.nanterre.francetelecom.fr> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.2.3.120616 x-originating-ip: [10.98.130.212] Content-Type: text/plain; charset="iso-8859-1" Content-ID: Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 25 Jun 2013 11:27:35 -0000 Hi Med, Yes agreed they are separate work items and should absolutely be worked on as part of this overall effort. On 6/25/13 2:19 AM, "mohamed.boucadair@orange.com" wrote: >Hi Jim, > >I think we are sync on the required protocol extensions and the merits of >the approach. > >On important aspect to consider is provisioning and policy enforcement: >as you can see in draft-boucadair-*, we separated these aspects from the >actual handling of packets and the behavior of involved nodes. I would >vote to include the work on the Information Model (data model may not be >in scope in a first step since DMs will depend on the favorite >provisioning protocol a network is using). > >Cheers, >Med > >>-----Message d'origine----- >>De : Jim Guichard (jguichar) [mailto:jguichar@cisco.com] >>Envoy=E9 : lundi 24 juin 2013 16:36 >>=C0 : BOUCADAIR Mohamed OLNC/OLN; adrian@olddog.co.uk; nsc@ietf.org >>Objet : Re: [nsc] What is a service? Why is NSC not SC? >> >>Hi Med, >> >>One further thought. All of the points that you make are well taken and >>honestly reflect one of the reasons why we have submitted our NSH >>proposal. Trying to define the characteristics and operation of each and >>every service function (NLF in your terminology) is imho looking at the >>problem from the wrong direction. What we want to be able to do is take >>the intelligence inherent within the network and present that information >>to whatever application/function is running within a service function so >>that it may take advantage of that data in whichever way it chooses. This >>requires that said intelligence is presented in a standardized format so >>that the developer of the service function can be sure of what >>information >>it can obtain from the network and how it might process that to provide >>the function it is designed to perform. It also provides the ability for >>one service function to interoperate with other service functions so as >>to >>build service chains. >> >>On 6/21/13 3:45 AM, "mohamed.boucadair@orange.com" >> wrote: >> >>>Hi Jim, all, >>> >>>I would even say: these functions are seen as black-boxes and no >>>assumption is to be made on the outcome of such functions. >>> >>>We tried to capture this and other assumptions in draft-boucadair-* >>> >>> o NLFs are handled as black-boxes; the technical characterization of >>> each NLF is not required. >>> o Not all NLFs can be characterized with a standard definition in >>> terms of technical description, detailed specification, etc. >>> o There is no global nor standard list of NLFs enabled in a given >>> administrative domain. The set of NLFs varies on a per deployment >>> basis. >>> o There is no global nor standard NLF chaining logic. The ordered >>> set of NLFs that need to be activated to deliver a given >>> connectivity service is specific to each administrative entity. >>> o The chaining of NLFs and the criteria to invoke some of them are >>> local to each administrative entity that operates a connectivity >>> network (also called administrative domain). >>> o NLF chaining logic and related policies should not be exposed >>> outside a given administrative domain. >>> o Several NLF chaining logics can be simultaneously enforced within >>> an administrative domain to meet business requirements. >>> >>>Cheers, >>>Med >>> >>>>-----Message d'origine----- >>>>De : nsc-bounces@ietf.org [mailto:nsc-bounces@ietf.org] De la part de >>>>Jim >>>>Guichard (jguichar) >>>>Envoy=E9 : jeudi 20 juin 2013 19:23 >>>>=C0 : adrian@olddog.co.uk; nsc@ietf.org >>>>Objet : Re: [nsc] What is a service? Why is NSC not SC? >>>> >>>>Hi Adrian, >>>> >>>>On 6/20/13 10:26 AM, "Adrian Farrel" wrote: >>>> >>>>>Thanks to Ken and Med for replies. >>>>> >>>>>Choosing to respond on Med's thread because he gave... >>>>> >>>>>> o NLF (Network-Located Function): refers to a function which is >>>>>> enabled in the network operated by an administrative entity. >>>>>>NLF >>>>>> can be for example: firewall (e.g., [RFC6092]), DPI (Deep >>>>>>Packet >>>>>> Inspection), LI (Lawful Intercept) module, NAT44 [RFC3022], >>>>>>NAT64 >>>>>> [RFC6146], DS-Lite AFTR [RFC6333], NPTv6 [RFC6296], HOST_ID >>>>>> injection, HTTP Header Enrichment function, TCP optimizer, >>>>>>etc. >>>>> >>>>>Something I pick out of this is that all the example functions are >>>>>packet-oriented functions (or possibly flow-oriented) and not >>>>>application >>>>>layer >>>>>services. I think this is significant and not accidental :-) >>>> >>>>Jim> actually the intent was that both packet-oriented and application >>>>layer service functions are within scope. It is certainly possible to >>>>envisage service chains that incorporate both of these types of >>>>functions >>>>and we should not restrict our work to one or the other but rather let >>>>the >>>>use cases dictate. >>>> >>>> >>>>>Can we capture that idea in a definition? >>>>> >>>>>I note that it is hinted at in Paul and Jim's BoF proposal text and it >>>>>is >>>>>there >>>>>because I asked them why this proposal was aimed at the Routing Area >>>>>not >>>>>the >>>>>Applications Area. To which the answer was "because we want to steer >>>>>packets or >>>>>flows to the place where the next service will be applied." >>>>> >>>>>Adrian >>>>> >>>>>_______________________________________________ >>>>>nsc mailing list >>>>>nsc@ietf.org >>>>>https://www.ietf.org/mailman/listinfo/nsc >>>> >>>>_______________________________________________ >>>>nsc mailing list >>>>nsc@ietf.org >>>>https://www.ietf.org/mailman/listinfo/nsc > From Ron_Parker@affirmednetworks.com Tue Jun 25 14:40:33 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C804311E8169 for ; Tue, 25 Jun 2013 14:40:33 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.598 X-Spam-Level: X-Spam-Status: No, score=-2.598 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HTML_MESSAGE=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id q4p+Ngw0mnOX for ; Tue, 25 Jun 2013 14:40:27 -0700 (PDT) Received: from hub021-ca-7.exch021.serverdata.net (hub021-ca-7.exch021.serverdata.net [64.78.56.72]) by ietfa.amsl.com (Postfix) with ESMTP id 3931111E8159 for ; Tue, 25 Jun 2013 14:40:22 -0700 (PDT) Received: from MBX021-W3-CA-2.exch021.domain.local ([10.254.4.78]) by HUB021-CA-7.exch021.domain.local ([10.254.4.109]) with mapi id 14.03.0123.003; Tue, 25 Jun 2013 14:40:16 -0700 From: Ron Parker To: "nsc@ietf.org" Thread-Topic: Comments on draft-boucadair-network-function-chaining-00 Thread-Index: Ac5x6/wxcUX+ReJaT8eNlB4kpr2L6Q== Date: Tue, 25 Jun 2013 21:40:16 +0000 Message-ID: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [98.110.150.224] Content-Type: multipart/alternative; boundary="_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66922DMBX021W3CA2exch_" MIME-Version: 1.0 Subject: [nsc] Comments on draft-boucadair-network-function-chaining-00 X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 25 Jun 2013 21:40:33 -0000 --_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66922DMBX021W3CA2exch_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Med, Christian, Thanks for the excellent document and to all that are contributing to this = worthwhile effort. I have included specific comments below and would be p= leased to see them spur some discussion on the thread. Best regards, Ron Parker Affirmed Networks 1.3 Objectives Is it desirable to support unidirectional chains or are only bidirectional = chains considered? For bidirectional chains, is NLF chronology always rev= ersed for each direction, or is it arbitrarily controlled? Are NLF's identified by IP address, FQDN, or both? Is it an objective for NLFC to explicitly handle load balancing towards mul= tiply-instantiated NLF's , or is load balancing in this regard considered t= o be part of each NLF? Is it an objective for NLFC to determine the availability of an NLF? If s= o, are there policies regarding what to do when an NLF becomes unavailable? 1.4 Assumptions All NLFs are reachable via standard IP routing techniques (i.e., no assumpt= ions regarding non-IP SDN protocols or techniques)? 1.5 Rationale Suggest comment about fragmentation be moved to Objectives section. 1.6 Generic Requirements Whether path stretch occurs or not is an administrative matter, assuming re= asonably correct IP routing topologies. 3.3 Building NLF Maps Does an NLF map specify the set of NLF's, only, or does it specify the set = of NLF's and the policy to be applied by each. For example, assume that F= irewall-A and NAT64-A are distinct instances of NLFs. Is {Firewall-A-prof= ile-1, NAT64-A-profile-1} distinct from {Firewall-A-profile-2, Nat64-A-prof= ile-1}. Or, alternatively, is the policy conveyed outside of the map? 4.1 NLFC Boundary Node I think the definition of ingress/egress here is problematic as it seems to= be implying topological direction (i.e., uplink and downlink in a mobile e= nvironment). If, however, the NLFC domain is thought of as residing in be= tween 2 non-NLFC domains (or segments of the same domain), then there is bo= th ingress and egress at both ends of the chain. A more descriptive appro= ach may be to use terms such as "West-ingress", "East-egress", "East-ingres= s", "West-egress". In a wireless domain, perhaps "north" and "south" wou= ld be more intuitive, but I think the connotation is more limiting in a gen= eral sense. 4.3 NLFC Ingress Node As an ingress node which connects to the non NLFC domain, when would an NLF= C Map Index be present? Also, this implies an encapsulation technique tha= t has not been previously introduced in the document. 5.1.1 NLFC Map Index If the individual entries in an NLFC Map indicate only the identity of the = NLF, then 8 or 16 bits to identify which NLFC Map would be sufficient. If= , however, the entries in an NLFC Map indicate NLF identity+policy, then 16= or 32 bits would be more appropriate. 5.1.3 Where to store NLFC Map Index in a packet? Consider usage of GRE key. Advantage is that in many cases, encapsulation= is required and GRE is a standard encapsulation technique which provides a= 32-bit general purpose field. Disadvantage is that some systems may requ= ire each unique GRE key to be provisioned as an IP interface. --_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66922DMBX021W3CA2exch_ Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable

Med, Christian,

 

Thanks for the excellent document and to all that ar= e contributing to this worthwhile effort.   I have included speci= fic comments below and would be pleased to see them spur some discussion on= the thread.

 

Best regards,

 

Ron Parker

Affirmed Networks

 

 

 

1.3 Objectives

Is it desirable to support unidirectional chains or = are only bidirectional chains considered?   For bidirectional cha= ins, is NLF chronology always reversed for each direction, or is it arbitra= rily  controlled?

Are NLF’s identified by IP address, FQDN, or b= oth?

Is it an objective for NLFC to explicitly handle loa= d balancing towards multiply-instantiated NLF’s , or is load balancin= g in this regard considered to be part of each NLF?

Is it an objective for NLFC to determine the availab= ility of an NLF?   If so, are there policies regarding what to do= when an NLF becomes unavailable?

 

1.4 Assumptions

All NLFs are reachable via standard IP routing techn= iques (i.e., no assumptions regarding non-IP SDN protocols or techniques)?<= o:p>

 

1.5 Rationale

Suggest comment about fragmentation be moved to Obje= ctives section.

 

1.6 Generic Requirements

Whether path stretch occurs or not is an administrat= ive matter, assuming reasonably correct IP routing topologies.

 

3.3 Building NLF Maps

Does an NLF map specify the set of NLF’s, only= , or does it specify the set of NLF’s and the policy to be applied by= each.   For example, assume that Firewall-A and NAT64-A are dist= inct instances of NLFs.   Is {Firewall-A-profile-1, NAT64-A-profi= le-1} distinct from {Firewall-A-profile-2, Nat64-A-profile-1}.   Or, a= lternatively, is the policy conveyed outside of the map?

 

4.1  NLFC Boundary Node

I think the definition of ingress/egress here is pro= blematic as it seems to be implying topological direction (i.e., uplink and= downlink in a mobile environment).   If, however, the NLFC domai= n is thought of as residing in between 2 non-NLFC domains (or segments of the same domain), then there is both ingress and e= gress at both ends of the chain.   A more descriptive approach ma= y be to use terms such as “West-ingress”, “East-egress= 221;, “East-ingress”, “West-egress”.  &nb= sp; In a wireless domain, perhaps “north” and “south” would be more intuitive, but I= think the connotation is more limiting in a general sense.

 

4.3 NLFC Ingress Node

As an ingress node which connects to the non NLFC do= main, when would an NLFC Map Index be present?   Also, this impli= es an encapsulation technique that has not been previously introduced in th= e document.

 

5.1.1 NLFC Map Index

If the individual entries in an NLFC Map indicate on= ly the identity of the NLF, then 8 or 16 bits to identify which NLFC Map wo= uld be sufficient.   If, however, the entries in an NLFC Map indi= cate NLF identity+policy, then 16 or 32 bits would be more appropriate.

 

5.1.3 Where to store NLFC Map Index in a packet?

Consider usage of GRE key.   Advantage is = that in many cases, encapsulation is required and GRE is a standard encapsu= lation technique which provides a 32-bit general purpose field.  = Disadvantage is that some systems may require each unique GRE key to be provisioned as an IP interface.

 

--_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66922DMBX021W3CA2exch_-- From hongyu.li@huawei.com Tue Jun 25 19:23:05 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id A3A1211E80F3 for ; Tue, 25 Jun 2013 19:23:04 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 5yQbI7fQnuDn for ; Tue, 25 Jun 2013 19:22:59 -0700 (PDT) Received: from lhrrgout.huawei.com (lhrrgout.huawei.com [194.213.3.17]) by ietfa.amsl.com (Postfix) with ESMTP id A0C5411E80BA for ; Tue, 25 Jun 2013 19:22:57 -0700 (PDT) Received: from 172.18.7.190 (EHLO lhreml203-edg.china.huawei.com) ([172.18.7.190]) by lhrrg01-dlp.huawei.com (MOS 4.3.5-GA FastPath queued) with ESMTP id AUK02793; Wed, 26 Jun 2013 02:22:50 +0000 (GMT) Received: from LHREML404-HUB.china.huawei.com (10.201.5.218) by lhreml203-edg.huawei.com (172.18.7.221) with Microsoft SMTP Server (TLS) id 14.1.323.7; Wed, 26 Jun 2013 03:21:58 +0100 Received: from SZXEML458-HUB.china.huawei.com (10.82.67.201) by lhreml404-hub.china.huawei.com (10.201.5.218) with Microsoft SMTP Server (TLS) id 14.1.323.7; Wed, 26 Jun 2013 10:22:48 +0800 Received: from szxeml560-mbx.china.huawei.com ([169.254.3.63]) by SZXEML458-HUB.china.huawei.com ([10.82.67.201]) with mapi id 14.01.0323.007; Wed, 26 Jun 2013 10:22:45 +0800 From: "Hongyu Li (Julio)" To: "Jim Guichard (jguichar)" , "Diego R. Lopez" Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrew//+rwICAADExAIAA8PYAgAUpkwCAACq9AIABLIuA//58R+A= Date: Wed, 26 Jun 2013 02:22:45 +0000 Message-ID: <6EB34CB5D82C4645B826C56144826EA97DF0ACFB@szxeml560-mbx.china.huawei.com> References: <68B171751455884590F8E38E96416F363F552B23@xmb-rcd-x01.cisco.com> In-Reply-To: <68B171751455884590F8E38E96416F363F552B23@xmb-rcd-x01.cisco.com> Accept-Language: en-US, zh-CN Content-Language: zh-CN X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [10.66.114.234] Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: base64 MIME-Version: 1.0 X-CFilter-Loop: Reflected Cc: "nsc@ietf.org" , "adrian@olddog.co.uk" , "mohamed.boucadair@orange.com" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 02:23:05 -0000 SGksDQoNCkN1cnJlbnQgdHdvIHR5cGVzIG9mIFNGIGlzIG9uZSB3YXkgb2YgY2xhc3NpZnlpbmcs IG1lYW5pbmcgU0YgaGF2aW5nIG9yIG5vdCBoYXZpbmcgdGhlIGNhcGFiaWxpdHkgb2YgZGVhbGlu ZyB3aXRoIGEgTlNDIGhlYWRlci4gSGVyZSwgSSdkIGxpa2UgdG8gaGlnaGxpZ2h0IHRoZXJlIGlz IGFub3RoZXIgd2F5IG9mIGNsYXNzaWZ5aW5nIFNGLCB3aGljaCB3ZSBuZWVkIHRvIHRha2UgaW50 byBjb25zaWRlcmF0aW9uLiBUaGF0IGlzLCB0aGUgU0YgbWF5IGJlIHRyYW5zcGFyZW50IG9yIHRl cm1pbmF0aW5nIHRvIGEgZmxvdy4gRm9yIGV4YW1wbGUsIGEgZmxvdyB3aWxsIGdvIHRocm91Z2gg YSBsb2FkIGJhbGFuY2VyIHRyYW5zcGFyZW50bHksIGJ1dCB3aWxsIGJlIHRlcm1pbmF0ZWQgYXQg YSBjYWNoZS4gVGhlc2UgdHdvIHR5cGVzIHdpbGwgaW5mbHVlbmNlIGhvdyBhIGNoYWluIGlzIG9y Z2FuaXplZCBhbmQgcmVxdWlyZW1lbnRzIG9mIHRoZSBwcm90b2NvbC4NCg0KTXkgdHdvIGNlbnRz Lg0KDQpIb25neXUNCg0KLS0tLS1PcmlnaW5hbCBNZXNzYWdlLS0tLS0NCkZyb206IG5zYy1ib3Vu Y2VzQGlldGYub3JnIFttYWlsdG86bnNjLWJvdW5jZXNAaWV0Zi5vcmddIE9uIEJlaGFsZiBPZiBK aW0gR3VpY2hhcmQgKGpndWljaGFyKQ0KU2VudDogVHVlc2RheSwgSnVuZSAyNSwgMjAxMyA3OjA0 IFBNDQpUbzogRGllZ28gUi4gTG9wZXoNCkNjOiBuc2NAaWV0Zi5vcmc7IGFkcmlhbkBvbGRkb2cu Y28udWs7IG1vaGFtZWQuYm91Y2FkYWlyQG9yYW5nZS5jb20NClN1YmplY3Q6IFJlOiBbbnNjXSBX aGF0IGlzIGEgc2VydmljZT8gV2h5IGlzIE5TQyBub3QgU0M/DQoNCkhpIERpZWdvLA0KDQpPbiA2 LzI0LzEzIDE6MDggUE0sICJEaWVnbyBSLiBMb3BleiIgPGRpZWdvQHRpZC5lcz4gd3JvdGU6DQoN Cj5IaSBKaW0sDQo+DQo+T24gMjQgSnVuIDIwMTMsIGF0IDE2OjM1ICwgSmltIEd1aWNoYXJkIChq Z3VpY2hhcikgd3JvdGU6DQo+Pg0KPj4gT25lIGZ1cnRoZXIgdGhvdWdodC4gQWxsIG9mIHRoZSBw b2ludHMgdGhhdCB5b3UgbWFrZSBhcmUgd2VsbCB0YWtlbiBhbmQNCj4+IGhvbmVzdGx5IHJlZmxl Y3Qgb25lIG9mIHRoZSByZWFzb25zIHdoeSB3ZSBoYXZlIHN1Ym1pdHRlZCBvdXIgTlNIDQo+PiBw cm9wb3NhbC4gVHJ5aW5nIHRvIGRlZmluZSB0aGUgY2hhcmFjdGVyaXN0aWNzIGFuZCBvcGVyYXRp b24gb2YgZWFjaCBhbmQNCj4+IGV2ZXJ5IHNlcnZpY2UgZnVuY3Rpb24gKE5MRiBpbiB5b3VyIHRl cm1pbm9sb2d5KSBpcyBpbWhvIGxvb2tpbmcgYXQgdGhlDQo+PiBwcm9ibGVtIGZyb20gdGhlIHdy b25nIGRpcmVjdGlvbi4gV2hhdCB3ZSB3YW50IHRvIGJlIGFibGUgdG8gZG8gaXMgdGFrZQ0KPj4g dGhlIGludGVsbGlnZW5jZSBpbmhlcmVudCB3aXRoaW4gdGhlIG5ldHdvcmsgYW5kIHByZXNlbnQg dGhhdA0KPj5pbmZvcm1hdGlvbg0KPj4gdG8gd2hhdGV2ZXIgYXBwbGljYXRpb24vZnVuY3Rpb24g aXMgcnVubmluZyB3aXRoaW4gYSBzZXJ2aWNlIGZ1bmN0aW9uIHNvDQo+PiB0aGF0IGl0IG1heSB0 YWtlIGFkdmFudGFnZSBvZiB0aGF0IGRhdGEgaW4gd2hpY2hldmVyIHdheSBpdCBjaG9vc2VzLg0K Pj5UaGlzDQo+PiByZXF1aXJlcyB0aGF0IHNhaWQgaW50ZWxsaWdlbmNlIGlzIHByZXNlbnRlZCBp biBhIHN0YW5kYXJkaXplZCBmb3JtYXQgc28NCj4+IHRoYXQgdGhlIGRldmVsb3BlciBvZiB0aGUg c2VydmljZSBmdW5jdGlvbiBjYW4gYmUgc3VyZSBvZiB3aGF0DQo+PmluZm9ybWF0aW9uDQo+PiBp dCBjYW4gb2J0YWluIGZyb20gdGhlIG5ldHdvcmsgYW5kIGhvdyBpdCBtaWdodCBwcm9jZXNzIHRo YXQgdG8gcHJvdmlkZQ0KPj4gdGhlIGZ1bmN0aW9uIGl0IGlzIGRlc2lnbmVkIHRvIHBlcmZvcm0u DQo+DQo+U28gdGhpcyByZXF1aXJlcyBhIHN0YW5kYXJkaXplZCBuZXR3b3JrLXRvLVNGIGludGVy ZmFjZS4gSSBhc3N1bWUgeW91DQo+cmVmZXIgaGVyZSB0byBhIG1lY2hhbmlzbSBieSB3aGljaCB0 aGUgU0YgY291bGQgbGVhcm4gd2hhdCBkYXRhIGl0IGNhbg0KPnJldHJpZXZlIGZyb20gdGhlIG5l dHdvcmsgb25jZSBkZXBsb3llZCBpbiBhIGNlcnRhaW4gbG9jYXRpb24sIGFzIHdlbGwgYXMNCj5k ZWZpbmluZyB0aGUgbWVjaGFuaXNtcyBmb3IgYXR0YWNoaW5nIGEgc2VydmljZSB0byB0aGF0IGxv Y2F0aW9uICh3aGV0aGVyDQo+YnkgbWVhbnMgb2YgYW4gYWRkcmVzcy9wb3J0IHBhaXIsIGEgVVJM LCBhIHR1bm5lbCBlbmRwb2ludC4uLikNCg0KSmltPiB5ZXMuDQoNCj4NCj4+IEl0IGFsc28gcHJv dmlkZXMgdGhlIGFiaWxpdHkgZm9yIG9uZSBzZXJ2aWNlIGZ1bmN0aW9uIHRvIGludGVyb3BlcmF0 ZQ0KPj53aXRoDQo+PiBvdGhlciBzZXJ2aWNlIGZ1bmN0aW9ucyBzbyBhcyB0byBidWlsZCBzZXJ2 aWNlIGNoYWlucy4NCj4NCj5BbmQgdGhpcyByZXF1aXJlcyBhIHN0YW5kYXJkaXplZCBTRiBtZXRh ZGF0YSBpbnRlcmZhY2UsIHNvIGVhY2ggU0YgY2FuDQo+YW5ub3VuY2Ugd2hhdCBpdCBleHBlY3Rz IGFzIGlucHV0IGFuZCBwcm92aWRlcyBhcyBvdXRwdXQsIHNvIGFuDQo+b3JjaGVzdHJhdG9yIGNh biBzdGF0aWNhbGx5IG9yIGV2ZW4gZHluYW1pY2FsbHkgYnVpbGQgdGhlIGFwcHJvcHJpYXRlDQo+ Y2hhaW4uIEFuZCB0aGlzIG9wZW4gdGhlIHF1ZXN0aW9uIG9uIHdoZXRoZXIgY2hhaW5pbmcgaXMg bGluZWFyIGFuZA0KPnVuaWRpcmVjdGlvbmFsLCBzbyB3ZSBjb3VsZCBzcGVhayBvZiBwaXBlbGlu aW5nLCBvciBpdCBpcyBmb3Jlc2VlYWJsZQ0KPm1vcmUgY29tcGxleCBjb21wb3NpdGlvbiBzdHJ1 Y3R1cmVzLg0KDQpKaW0+IHllcyBhbHRob3VnaCB0aGUgU0YgbWF5IG9yIG1heSBub3QgYW5ub3Vu Y2UgaXRzIGNhcGFiaWxpdGllcyAoaW4gdGhlDQpsYXR0ZXIgY2FzZSB0aGUgb3JjaGVzdHJhdG9y IHdvdWxkIGJlIHRvbGQgZGlyZWN0bHkpLiBJbiB0ZXJtcyBvZiBzZXJ2aWNlDQpjaGFpbnMgd2Ug aGF2ZSB0byB3b3JrIHRocm91Z2ggYm90aCB1bmlkaXJlY3Rpb25hbCBhbmQgYmlkaXJlY3Rpb25h bCBhcw0Kd2VsbCBhcyBjb21wbGV4IGNoYWlucyAod2hlcmUgdHJhZmZpYyBtYXkgYmUgZGl2ZXJ0 ZWQgbWlkLWNoYWluIGZvcg0KZXhhbXBsZSBvbiB0byBhIGRpZmZlcmVudCBjaGFpbikuDQoNCj4N Cj5CZSBnb29kZSwNCj4NCj4NCj4tLQ0KPiJFc3RhIHZleiBubyBmYWxsYXJlbW9zLCBEb2N0b3Ig SW5maWVybm8iDQo+DQo+RHIgRGllZ28gUi4gTG9wZXoNCj5UZWxlZm9uaWNhIEkrRA0KPmh0dHA6 Ly9wZW9wbGUudGlkLmVzL2RpZWdvLmxvcGV6Lw0KPg0KPmUtbWFpbDogZGllZ29AdGlkLmVzDQo+ VGVsOiAgICArMzQgOTEzIDEyOSAwNDENCj5Nb2JpbGU6ICszNCA2ODIgMDUxIDA5MQ0KPi0tLS0t LS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tDQo+DQo+DQo+X19fX19fX19fX19f X19fX19fX19fX19fX19fX19fX18NCj4NCj5Fc3RlIG1lbnNhamUgc2UgZGlyaWdlIGV4Y2x1c2l2 YW1lbnRlIGEgc3UgZGVzdGluYXRhcmlvLiBQdWVkZSBjb25zdWx0YXINCj5udWVzdHJhIHBvbMOt dGljYSBkZSBlbnbDrW8geSByZWNlcGNpw7NuIGRlIGNvcnJlbyBlbGVjdHLDs25pY28gZW4gZWwg ZW5sYWNlDQo+c2l0dWFkbyBtw6FzIGFiYWpvLg0KPlRoaXMgbWVzc2FnZSBpcyBpbnRlbmRlZCBl eGNsdXNpdmVseSBmb3IgaXRzIGFkZHJlc3NlZS4gV2Ugb25seSBzZW5kIGFuZA0KPnJlY2VpdmUg ZW1haWwgb24gdGhlIGJhc2lzIG9mIHRoZSB0ZXJtcyBzZXQgb3V0IGF0Og0KPmh0dHA6Ly93d3cu dGlkLmVzL0VTL1BBR0lOQVMvZGlzY2xhaW1lci5hc3B4DQoNCl9fX19fX19fX19fX19fX19fX19f X19fX19fX19fX19fX19fX19fX19fX19fX19fDQpuc2MgbWFpbGluZyBsaXN0DQpuc2NAaWV0Zi5v cmcNCmh0dHBzOi8vd3d3LmlldGYub3JnL21haWxtYW4vbGlzdGluZm8vbnNjDQo= From hongyu.li@huawei.com Tue Jun 25 19:51:25 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C948011E8191 for ; Tue, 25 Jun 2013 19:51:25 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Gk16AE5-5AhL for ; Tue, 25 Jun 2013 19:51:21 -0700 (PDT) Received: from lhrrgout.huawei.com (lhrrgout.huawei.com [194.213.3.17]) by ietfa.amsl.com (Postfix) with ESMTP id 356D411E80F8 for ; Tue, 25 Jun 2013 19:51:21 -0700 (PDT) Received: from 172.18.7.190 (EHLO lhreml203-edg.china.huawei.com) ([172.18.7.190]) by lhrrg01-dlp.huawei.com (MOS 4.3.5-GA FastPath queued) with ESMTP id AUK04654; Wed, 26 Jun 2013 02:51:19 +0000 (GMT) Received: from LHREML406-HUB.china.huawei.com (10.201.5.243) by lhreml203-edg.huawei.com (172.18.7.221) with Microsoft SMTP Server (TLS) id 14.1.323.7; Wed, 26 Jun 2013 03:50:25 +0100 Received: from SZXEML406-HUB.china.huawei.com (10.82.67.93) by lhreml406-hub.china.huawei.com (10.201.5.243) with Microsoft SMTP Server (TLS) id 14.1.323.7; Wed, 26 Jun 2013 03:51:16 +0100 Received: from szxeml560-mbx.china.huawei.com ([169.254.3.63]) by szxeml406-hub.china.huawei.com ([10.82.67.93]) with mapi id 14.01.0323.007; Wed, 26 Jun 2013 10:51:11 +0800 From: "Hongyu Li (Julio)" To: "Diego R. Lopez" , " " Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrew//+rwICAADExAIAA8PYAgAQjwYD//Bu+0A== Date: Wed, 26 Jun 2013 02:51:10 +0000 Message-ID: <6EB34CB5D82C4645B826C56144826EA97DF0AD29@szxeml560-mbx.china.huawei.com> References: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> In-Reply-To: Accept-Language: en-US, zh-CN Content-Language: zh-CN X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [10.66.114.234] Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: base64 MIME-Version: 1.0 X-CFilter-Loop: Reflected Cc: "nsc@ietf.org" , "Jim Guichard \(jguichar\)" , "adrian@olddog.co.uk" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 02:51:25 -0000 SGksDQoNClBsZWFzZSBhbGxvdyBtZSB0byBjaGlwIGluIG9uIHRoZSBvcmlnaW5hbCBxdWVzdGlv bnMuDQoNCk5TQyBtYXkgYmUgaW50ZXJwcmV0ZWQgaW4gdHdvIHdheXMuIE9uZSBpcyB3aGF0J3Mg Y2hhaW5lZCBhcmUgbmV0d29yayBmdW5jdGlvbnMuIFRoZSBvdGhlciBpcyBzZXJ2aWNlIGZ1bmN0 aW9ucyBhcmUgY2hhaW5lZCBvbiBuZXR3b3JrIGxheWVyLiBCb3RoIHNob3VsZCBiZSBpbiB0aGUg c2NvcGUuIEluIG90aGVyIHdvcmRzLCB3ZSBhcmUgc2Vla2luZyBjaGFpbmluZyBTRnMgb24gbmV0 d29yayBsZXZlbCwgd2hpbGUgU0YgY291bGQgYmUgbmV0d29yayBzZXJ2aWNlIGZ1bmN0aW9ucyBv ciBoaWdoZXIgbGF5ZXIgc2VydmljZSBmdW5jdGlvbnMuDQoNCkFub3RoZXIgcG9pbnQgSSB3YW5u YSBtYWtlIGlzIHRoYXQgSU1ITywgb24ga2V5IGNoYXJhY3RlciBvZiBOU0MgaXMgdHJhZmZpYyBp cyBmb3J3YXJkZWQgcGVyIHNlcnZpY2UgY2hhcmFjdGVyaXN0aWNzIG90aGVyIHRoYW4gZGVzdGlu YXRpb24gYWRkcmVzc2VzLiBUaGlzIGlzIHRoZSBrZXkgdmFsdWUgTlNDIFdHLCBpZiB3aWxsIGJl IGZvcm1lZCwgc2hvdWxkIHdvcmsgb24uIFNlcnZpY2UgY2hhcmFjdGVyaXN0aWNzIG1heSBpbmNs dWRlIHRoZSBuZXR3b3JrIHNlcnZpY2UgYSB1c2UgaGFzIHN1YnNjcmliZWQgdG8sIGUuZy4gcGFy ZW50YWwgY29udHJvbCwgZmlyZXdhbGwuIFRoYXQgbWVhbnMgYSBwYWNrZXQgaXMgZm9yd2FyZGVk IGludG8gYSBzZXJ2aWNlIGNoYWluLCB3aGVuIGEgY2xhc3NpZmllciBtYXRjaGVzIHNvbWUgcnVs ZXMgb2Ygc2VydmljZSBjaGFyYWN0ZXJpc3RpY3MsIHdoaWNoIG1heSBiZSByZWZsZWN0ZWQgYnkg YWNjZXNzIHBvcnRzLCBvciBzZWdtZW50cyBvciB0dXBsZXMgaW4gYSBwYWNrZXQuDQoNCkNoZWVy cywNCkhvbmd5dQ0KDQotLS0tLU9yaWdpbmFsIE1lc3NhZ2UtLS0tLQ0KRnJvbTogbnNjLWJvdW5j ZXNAaWV0Zi5vcmcgW21haWx0bzpuc2MtYm91bmNlc0BpZXRmLm9yZ10gT24gQmVoYWxmIE9mIERp ZWdvIFIuIExvcGV6DQpTZW50OiBNb25kYXksIEp1bmUgMjQsIDIwMTMgNjo1OCBBTQ0KVG86IDxt b2hhbWVkLmJvdWNhZGFpckBvcmFuZ2UuY29tPiA8bW9oYW1lZC5ib3VjYWRhaXJAb3JhbmdlLmNv bT4NCkNjOiBuc2NAaWV0Zi5vcmc7IEppbSBHdWljaGFyZCAoamd1aWNoYXIpOyBhZHJpYW5Ab2xk ZG9nLmNvLnVrDQpTdWJqZWN0OiBSZTogW25zY10gV2hhdCBpcyBhIHNlcnZpY2U/IFdoeSBpcyBO U0Mgbm90IFNDPw0KDQpIaSBhbGwsDQoNCkxvb2tpbmcgYXQgdGhpcyB0aHJlYWQsIGNvbWJpbmlu ZyB0aGUgZGlzdGluY3Rpb24gYmV0d2VlbiAiZnVuY3Rpb24iIGFuZCAic2VydmljZSIgcHJvdmlk ZWQgYnkgQmVuc29uLCBhbmQgdGhlIGFzc3VtcHRpb25zIE1lZCBsaXN0cyBpbiB3aGF0IHJlbGF0 ZXMgdG8gaG93IHBvbGljeSBhcHBsaWVzIHRvIHRoZW0gd2l0aGluIGFuIGFkbWluaXN0cmF0aXZl IGRvbWFpbiwgSSBhbSBpbiBmYXZvciBvZiB0YWxraW5nIG9mIGZ1bmN0aW9ucyB0aGF0IGFyZSBj b21iaW5lZCAoY2hhaW5lZCkgdG8gYnVpbGQgc2VydmljZXMsIGF0IGFueSByZWN1cnNpb24gbGV2 ZWw6IGEgc2VydmljZSBjYW4gYmUgY29tcG9zZWQgKGNoYWluZWQpIGFzIGEgZnVuY3Rpb24gdG8g YnVpbGQgYSBtb3JlIGNvbXBsZXggc2VydmljZS4NCg0KUHJvYmFibHkgdGhlIFNvbG9tb25pYyBw cm9wb3NhbCBmcm9tIEppbSBpcyB0aGUgb25lIG1ha2luZyBtb3JlIHNlbnNlLi4uDQoNCkJlIGdv b2RlLA0KDQpPbiAyMSBKdW4gMjAxMywgYXQgMDk6NDUgLCA8bW9oYW1lZC5ib3VjYWRhaXJAb3Jh bmdlLmNvbT4NCiA8bW9oYW1lZC5ib3VjYWRhaXJAb3JhbmdlLmNvbT4gd3JvdGU6DQoNCj4gSGkg SmltLCBhbGwsDQo+DQo+IEkgd291bGQgZXZlbiBzYXk6IHRoZXNlIGZ1bmN0aW9ucyBhcmUgc2Vl biBhcyBibGFjay1ib3hlcyBhbmQgbm8gYXNzdW1wdGlvbiBpcyB0byBiZSBtYWRlIG9uIHRoZSBv dXRjb21lIG9mIHN1Y2ggZnVuY3Rpb25zLg0KPg0KPiBXZSB0cmllZCB0byBjYXB0dXJlIHRoaXMg YW5kIG90aGVyIGFzc3VtcHRpb25zIGluIGRyYWZ0LWJvdWNhZGFpci0qDQo+DQo+ICAgbyAgTkxG cyBhcmUgaGFuZGxlZCBhcyBibGFjay1ib3hlczsgdGhlIHRlY2huaWNhbCBjaGFyYWN0ZXJpemF0 aW9uIG9mDQo+ICAgICAgZWFjaCBOTEYgaXMgbm90IHJlcXVpcmVkLg0KPiAgIG8gIE5vdCBhbGwg TkxGcyBjYW4gYmUgY2hhcmFjdGVyaXplZCB3aXRoIGEgc3RhbmRhcmQgZGVmaW5pdGlvbiBpbg0K PiAgICAgIHRlcm1zIG9mIHRlY2huaWNhbCBkZXNjcmlwdGlvbiwgZGV0YWlsZWQgc3BlY2lmaWNh dGlvbiwgZXRjLg0KPiAgIG8gIFRoZXJlIGlzIG5vIGdsb2JhbCBub3Igc3RhbmRhcmQgbGlzdCBv ZiBOTEZzIGVuYWJsZWQgaW4gYSBnaXZlbg0KPiAgICAgIGFkbWluaXN0cmF0aXZlIGRvbWFpbi4g IFRoZSBzZXQgb2YgTkxGcyB2YXJpZXMgb24gYSBwZXIgZGVwbG95bWVudA0KPiAgICAgIGJhc2lz Lg0KPiAgIG8gIFRoZXJlIGlzIG5vIGdsb2JhbCBub3Igc3RhbmRhcmQgTkxGIGNoYWluaW5nIGxv Z2ljLiAgVGhlIG9yZGVyZWQNCj4gICAgICBzZXQgb2YgTkxGcyB0aGF0IG5lZWQgdG8gYmUgYWN0 aXZhdGVkIHRvIGRlbGl2ZXIgYSBnaXZlbg0KPiAgICAgIGNvbm5lY3Rpdml0eSBzZXJ2aWNlIGlz IHNwZWNpZmljIHRvIGVhY2ggYWRtaW5pc3RyYXRpdmUgZW50aXR5Lg0KPiAgIG8gIFRoZSBjaGFp bmluZyBvZiBOTEZzIGFuZCB0aGUgY3JpdGVyaWEgdG8gaW52b2tlIHNvbWUgb2YgdGhlbSBhcmUN Cj4gICAgICBsb2NhbCB0byBlYWNoIGFkbWluaXN0cmF0aXZlIGVudGl0eSB0aGF0IG9wZXJhdGVz IGEgY29ubmVjdGl2aXR5DQo+ICAgICAgbmV0d29yayAoYWxzbyBjYWxsZWQgYWRtaW5pc3RyYXRp dmUgZG9tYWluKS4NCj4gICBvICBOTEYgY2hhaW5pbmcgbG9naWMgYW5kIHJlbGF0ZWQgcG9saWNp ZXMgc2hvdWxkIG5vdCBiZSBleHBvc2VkDQo+ICAgICAgb3V0c2lkZSBhIGdpdmVuIGFkbWluaXN0 cmF0aXZlIGRvbWFpbi4NCj4gICBvICBTZXZlcmFsIE5MRiBjaGFpbmluZyBsb2dpY3MgY2FuIGJl IHNpbXVsdGFuZW91c2x5IGVuZm9yY2VkIHdpdGhpbg0KPiAgICAgIGFuIGFkbWluaXN0cmF0aXZl IGRvbWFpbiB0byBtZWV0IGJ1c2luZXNzIHJlcXVpcmVtZW50cy4NCj4NCj4gQ2hlZXJzLA0KPiBN ZWQNCj4NCj4+IC0tLS0tTWVzc2FnZSBkJ29yaWdpbmUtLS0tLQ0KPj4gRGUgOiBuc2MtYm91bmNl c0BpZXRmLm9yZyBbbWFpbHRvOm5zYy1ib3VuY2VzQGlldGYub3JnXSBEZSBsYSBwYXJ0IGRlIEpp bQ0KPj4gR3VpY2hhcmQgKGpndWljaGFyKQ0KPj4gRW52b3nDqSA6IGpldWRpIDIwIGp1aW4gMjAx MyAxOToyMw0KPj4gw4AgOiBhZHJpYW5Ab2xkZG9nLmNvLnVrOyBuc2NAaWV0Zi5vcmcNCj4+IE9i amV0IDogUmU6IFtuc2NdIFdoYXQgaXMgYSBzZXJ2aWNlPyBXaHkgaXMgTlNDIG5vdCBTQz8NCj4+ DQo+PiBIaSBBZHJpYW4sDQo+Pg0KPj4gT24gNi8yMC8xMyAxMDoyNiBBTSwgIkFkcmlhbiBGYXJy ZWwiIDxhZHJpYW5Ab2xkZG9nLmNvLnVrPiB3cm90ZToNCj4+DQo+Pj4gVGhhbmtzIHRvIEtlbiBh bmQgTWVkIGZvciByZXBsaWVzLg0KPj4+DQo+Pj4gQ2hvb3NpbmcgdG8gcmVzcG9uZCBvbiBNZWQn cyB0aHJlYWQgYmVjYXVzZSBoZSBnYXZlLi4uDQo+Pj4NCj4+Pj4gICBvICBOTEYgKE5ldHdvcmst TG9jYXRlZCBGdW5jdGlvbik6IHJlZmVycyB0byBhIGZ1bmN0aW9uIHdoaWNoIGlzDQo+Pj4+ICAg ICAgZW5hYmxlZCBpbiB0aGUgbmV0d29yayBvcGVyYXRlZCBieSBhbiBhZG1pbmlzdHJhdGl2ZSBl bnRpdHkuICBOTEYNCj4+Pj4gICAgICBjYW4gYmUgZm9yIGV4YW1wbGU6IGZpcmV3YWxsIChlLmcu LCBbUkZDNjA5Ml0pLCBEUEkgKERlZXAgUGFja2V0DQo+Pj4+ICAgICAgSW5zcGVjdGlvbiksIExJ IChMYXdmdWwgSW50ZXJjZXB0KSBtb2R1bGUsIE5BVDQ0IFtSRkMzMDIyXSwgTkFUNjQNCj4+Pj4g ICAgICBbUkZDNjE0Nl0sIERTLUxpdGUgQUZUUiBbUkZDNjMzM10sIE5QVHY2IFtSRkM2Mjk2XSwg SE9TVF9JRA0KPj4+PiAgICAgIGluamVjdGlvbiwgSFRUUCBIZWFkZXIgRW5yaWNobWVudCBmdW5j dGlvbiwgVENQIG9wdGltaXplciwgZXRjLg0KPj4+DQo+Pj4gU29tZXRoaW5nIEkgcGljayBvdXQg b2YgdGhpcyBpcyB0aGF0IGFsbCB0aGUgZXhhbXBsZSBmdW5jdGlvbnMgYXJlDQo+Pj4gcGFja2V0 LW9yaWVudGVkIGZ1bmN0aW9ucyAob3IgcG9zc2libHkgZmxvdy1vcmllbnRlZCkgYW5kIG5vdCBh cHBsaWNhdGlvbg0KPj4+IGxheWVyDQo+Pj4gc2VydmljZXMuIEkgdGhpbmsgdGhpcyBpcyBzaWdu aWZpY2FudCBhbmQgbm90IGFjY2lkZW50YWwgOi0pDQo+Pg0KPj4gSmltPiBhY3R1YWxseSB0aGUg aW50ZW50IHdhcyB0aGF0IGJvdGggcGFja2V0LW9yaWVudGVkIGFuZCBhcHBsaWNhdGlvbg0KPj4g bGF5ZXIgc2VydmljZSBmdW5jdGlvbnMgYXJlIHdpdGhpbiBzY29wZS4gSXQgaXMgY2VydGFpbmx5 IHBvc3NpYmxlIHRvDQo+PiBlbnZpc2FnZSBzZXJ2aWNlIGNoYWlucyB0aGF0IGluY29ycG9yYXRl IGJvdGggb2YgdGhlc2UgdHlwZXMgb2YgZnVuY3Rpb25zDQo+PiBhbmQgd2Ugc2hvdWxkIG5vdCBy ZXN0cmljdCBvdXIgd29yayB0byBvbmUgb3IgdGhlIG90aGVyIGJ1dCByYXRoZXIgbGV0IHRoZQ0K Pj4gdXNlIGNhc2VzIGRpY3RhdGUuDQo+Pg0KPj4NCj4+PiBDYW4gd2UgY2FwdHVyZSB0aGF0IGlk ZWEgaW4gYSBkZWZpbml0aW9uPw0KPj4+DQo+Pj4gSSBub3RlIHRoYXQgaXQgaXMgaGludGVkIGF0 IGluIFBhdWwgYW5kIEppbSdzIEJvRiBwcm9wb3NhbCB0ZXh0IGFuZCBpdCBpcw0KPj4+IHRoZXJl DQo+Pj4gYmVjYXVzZSBJIGFza2VkIHRoZW0gd2h5IHRoaXMgcHJvcG9zYWwgd2FzIGFpbWVkIGF0 IHRoZSBSb3V0aW5nIEFyZWEgbm90DQo+Pj4gdGhlDQo+Pj4gQXBwbGljYXRpb25zIEFyZWEuIFRv IHdoaWNoIHRoZSBhbnN3ZXIgd2FzICJiZWNhdXNlIHdlIHdhbnQgdG8gc3RlZXINCj4+PiBwYWNr ZXRzIG9yDQo+Pj4gZmxvd3MgdG8gdGhlIHBsYWNlIHdoZXJlIHRoZSBuZXh0IHNlcnZpY2Ugd2ls bCBiZSBhcHBsaWVkLiINCj4+Pg0KPj4+IEFkcmlhbg0KPj4+DQo+Pj4gX19fX19fX19fX19fX19f X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX18NCj4+PiBuc2MgbWFpbGluZyBsaXN0DQo+ Pj4gbnNjQGlldGYub3JnDQo+Pj4gaHR0cHM6Ly93d3cuaWV0Zi5vcmcvbWFpbG1hbi9saXN0aW5m by9uc2MNCj4+DQo+PiBfX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19f X19fXw0KPj4gbnNjIG1haWxpbmcgbGlzdA0KPj4gbnNjQGlldGYub3JnDQo+PiBodHRwczovL3d3 dy5pZXRmLm9yZy9tYWlsbWFuL2xpc3RpbmZvL25zYw0KPiBfX19fX19fX19fX19fX19fX19fX19f X19fX19fX19fX19fX19fX19fX19fX19fXw0KPiBuc2MgbWFpbGluZyBsaXN0DQo+IG5zY0BpZXRm Lm9yZw0KPiBodHRwczovL3d3dy5pZXRmLm9yZy9tYWlsbWFuL2xpc3RpbmZvL25zYw0KDQoNCi0t DQoiRXN0YSB2ZXogbm8gZmFsbGFyZW1vcywgRG9jdG9yIEluZmllcm5vIg0KDQpEciBEaWVnbyBS LiBMb3Bleg0KVGVsZWZvbmljYSBJK0QNCmh0dHA6Ly9wZW9wbGUudGlkLmVzL2RpZWdvLmxvcGV6 Lw0KDQplLW1haWw6IGRpZWdvQHRpZC5lcw0KVGVsOiAgICArMzQgOTEzIDEyOSAwNDENCk1vYmls ZTogKzM0IDY4MiAwNTEgMDkxDQotLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0t LS0tLQ0KDQoNCl9fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fDQoNCkVzdGUgbWVuc2Fq ZSBzZSBkaXJpZ2UgZXhjbHVzaXZhbWVudGUgYSBzdSBkZXN0aW5hdGFyaW8uIFB1ZWRlIGNvbnN1 bHRhciBudWVzdHJhIHBvbMOtdGljYSBkZSBlbnbDrW8geSByZWNlcGNpw7NuIGRlIGNvcnJlbyBl bGVjdHLDs25pY28gZW4gZWwgZW5sYWNlIHNpdHVhZG8gbcOhcyBhYmFqby4NClRoaXMgbWVzc2Fn ZSBpcyBpbnRlbmRlZCBleGNsdXNpdmVseSBmb3IgaXRzIGFkZHJlc3NlZS4gV2Ugb25seSBzZW5k IGFuZCByZWNlaXZlIGVtYWlsIG9uIHRoZSBiYXNpcyBvZiB0aGUgdGVybXMgc2V0IG91dCBhdDoN Cmh0dHA6Ly93d3cudGlkLmVzL0VTL1BBR0lOQVMvZGlzY2xhaW1lci5hc3B4DQpfX19fX19fX19f X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fXw0KbnNjIG1haWxpbmcgbGlzdA0K bnNjQGlldGYub3JnDQpodHRwczovL3d3dy5pZXRmLm9yZy9tYWlsbWFuL2xpc3RpbmZvL25zYw0K From mohamed.boucadair@orange.com Tue Jun 25 23:31:44 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6F58E21F9F40 for ; Tue, 25 Jun 2013 23:31:44 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.247 X-Spam-Level: X-Spam-Status: No, score=-2.247 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HELO_EQ_FR=0.35, HTML_MESSAGE=0.001, UNPARSEABLE_RELAY=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Q59eC452TaKl for ; Tue, 25 Jun 2013 23:31:39 -0700 (PDT) Received: from relais-inet.francetelecom.com (relais-ias91.francetelecom.com [193.251.215.91]) by ietfa.amsl.com (Postfix) with ESMTP id 8365B21E8119 for ; Tue, 25 Jun 2013 23:31:38 -0700 (PDT) Received: from omfedm08.si.francetelecom.fr (unknown [xx.xx.xx.4]) by omfedm14.si.francetelecom.fr (ESMTP service) with ESMTP id 1DB8222D09E; Wed, 26 Jun 2013 08:31:37 +0200 (CEST) Received: from PUEXCH81.nanterre.francetelecom.fr (unknown [10.101.44.34]) by omfedm08.si.francetelecom.fr (ESMTP service) with ESMTP id E604423805E; Wed, 26 Jun 2013 08:31:36 +0200 (CEST) Received: from PUEXCB1B.nanterre.francetelecom.fr ([10.101.44.13]) by PUEXCH81.nanterre.francetelecom.fr ([10.101.44.34]) with mapi; Wed, 26 Jun 2013 08:31:36 +0200 From: To: Ron Parker , "nsc@ietf.org" Date: Wed, 26 Jun 2013 08:31:35 +0200 Thread-Topic: Comments on draft-boucadair-network-function-chaining-00 Thread-Index: Ac5x6/wxcUX+ReJaT8eNlB4kpr2L6QART59w Message-ID: <94C682931C08B048B7A8645303FDC9F36EDB5C7F8F@PUEXCB1B.nanterre.francetelecom.fr> References: In-Reply-To: Accept-Language: fr-FR Content-Language: fr-FR X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: fr-FR Content-Type: multipart/alternative; boundary="_000_94C682931C08B048B7A8645303FDC9F36EDB5C7F8FPUEXCB1Bnante_" MIME-Version: 1.0 X-PMX-Version: 5.6.1.2065439, Antispam-Engine: 2.7.2.376379, Antispam-Data: 2013.6.26.44224 Subject: Re: [nsc] Comments on draft-boucadair-network-function-chaining-00 X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 06:31:44 -0000 --_000_94C682931C08B048B7A8645303FDC9F36EDB5C7F8FPUEXCB1Bnante_ Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Dear Ron, Thank you for the review. Please see inline. Cheers, Med De : nsc-bounces@ietf.org [mailto:nsc-bounces@ietf.org] De la part de Ron P= arker Envoy=E9 : mardi 25 juin 2013 23:40 =C0 : nsc@ietf.org Objet : [nsc] Comments on draft-boucadair-network-function-chaining-00 Med, Christian, Thanks for the excellent document and to all that are contributing to this = worthwhile effort. I have included specific comments below and would be p= leased to see them spur some discussion on the thread. Best regards, Ron Parker Affirmed Networks 1.3 Objectives Is it desirable to support unidirectional chains or are only bidirectional = chains considered? [Med] The current text (section 3.3) indicates distinct chains can be appli= ed for inbound and outbound traffic. The direction of the chain is not incl= uded as an attribute of the NLFC Map but this is realized using two chains:= i.e., two NLFC Maps will be installed in the NLFC Policy Table. For bidirectional chains, is NLF chronology always reversed for each dire= ction, or is it arbitrarily controlled? [Med] The current text allows to specify distinct NLFC Map for each directi= on. Are NLF's identified by IP address, FQDN, or both? [Med] I guess your question is about locators not identifiers. The current = text says both can be supported but we recommend the use of an IP address s= ince no resolution capability would be needed in the NLF-enabled node. Reso= lution capabilities (and advanced TE functions) can be enabled only at the = PDP (Policy Decision Point); no need to overload NLF-enabled nodes with suc= h features. Is it an objective for NLFC to explicitly handle load balancing towards mul= tiply-instantiated NLF's , or is load balancing in this regard considered t= o be part of each NLF? [Med] Yes. The current text says: o Advanced features (e.g., load balancing objectives) are also policy-based. Invoked enforcement points are not aware of such objectives. Is it an objective for NLFC to determine the availability of an NLF? If s= o, are there policies regarding what to do when an NLF becomes unavailable? [Med] Good point. FYI, we are currently working on a OSPF extension for thi= s purpose. We will share the text soon. 1.4 Assumptions All NLFs are reachable via standard IP routing techniques (i.e., no assumpt= ions regarding non-IP SDN protocols or techniques)? [Med] How FIBs and RIBs are populated is orthogonal to the overall procedur= e. 1.5 Rationale Suggest comment about fragmentation be moved to Objectives section. [Med] I will see how to update the text. 1.6 Generic Requirements Whether path stretch occurs or not is an administrative matter, assuming re= asonably correct IP routing topologies. [Med] Agree. This is deployment-specific. 3.3 Building NLF Maps Does an NLF map specify the set of NLF's, only, or does it specify the set = of NLF's and the policy to be applied by each. For example, assume that F= irewall-A and NAT64-A are distinct instances of NLFs. Is {Firewall-A-prof= ile-1, NAT64-A-profile-1} distinct from {Firewall-A-profile-2, Nat64-A-prof= ile-1}. Or, alternatively, is the policy conveyed outside of the map? [Med] I'm not sure I get the point. What do you mean by "profile" here? 4.1 NLFC Boundary Node I think the definition of ingress/egress here is problematic as it seems to= be implying topological direction (i.e., uplink and downlink in a mobile e= nvironment). If, however, the NLFC domain is thought of as residing in be= tween 2 non-NLFC domains (or segments of the same domain), then there is bo= th ingress and egress at both ends of the chain. A more descriptive appro= ach may be to use terms such as "West-ingress", "East-egress", "East-ingres= s", "West-egress". In a wireless domain, perhaps "north" and "south" wou= ld be more intuitive, but I think the connotation is more limiting in a gen= eral sense. [Med] Each NLFC Boundary Node behaves in both the ingress and egress roles.= It behaves as "ingress" for traffic entering the domain and as "egress" fo= r traffic existing the domain. 4.3 NLFC Ingress Node As an ingress node which connects to the non NLFC domain, when would an NLF= C Map Index be present? Also, this implies an encapsulation technique tha= t has not been previously introduced in the document. [Med] Just like setting/re-setting DSCP values at the boundaries of a domai= n, the NLFC Map Index can be set by any "non-legitimate" node. This is even= a basic attack to disturb a service or try granting access to non-authoriz= ed services. This is why we recommend to enforce a basic policy to not trus= t any NLFC Map Index received from adjacent domains. 5.1.1 NLFC Map Index If the individual entries in an NLFC Map indicate only the identity of the = NLF, then 8 or 16 bits to identify which NLFC Map would be sufficient. If= , however, the entries in an NLFC Map indicate NLF identity+policy, then 16= or 32 bits would be more appropriate. [Med] The policy is not signaled in packets; only the index of the NLFC Map= is. 5.1.3 Where to store NLFC Map Index in a packet? Consider usage of GRE key. Advantage is that in many cases, encapsulation= is required and GRE is a standard encapsulation technique which provides a= 32-bit general purpose field. Disadvantage is that some systems may requ= ire each unique GRE key to be provisioned as an IP interface. [Med] Will add a new entry for it. Another inconvenient I see is for cases = where encapsulation is not needed at all. This is typically when the next N= LF node is an adjacent node (see Section 6.2). --_000_94C682931C08B048B7A8645303FDC9F36EDB5C7F8FPUEXCB1Bnante_ Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

Dear Ron,

 

Thank you for the review.

Please see inline.

 

= Cheers,

Med=

 

De : nsc-bounces@ietf.o= rg [mailto:nsc-bounces@ietf.org] De la part de Ron Parker
Envo= y=E9 : mardi 25 juin 2013 23:40
=C0 : nsc@ietf.org<= br>Objet : [nsc] Comments on draft-boucadair-network-function-c= haining-00

 = ;

Med, Christian,

 

Thanks for the excellent docu= ment and to all that are contributing to this worthwhile effort.  = ; I have included specific comments below and would be pleased to see them = spur some discussion on the thread.

 

Best regards,

 

Ron Parker

Affirmed Networks

 

 

 

1.3 Objecti= ves

Is it desi= rable to support unidirectional chains or are only bidirectional chains con= sidered? 

[Med] The current text (section 3.3) indicate= s distinct chains can be applied for inbound and outbound traffic. The dire= ction of the chain is not included as an attribute of the NLFC Map but this= is realized using two chains: i.e., two NLFC Maps will be installed in the= NLFC Policy Table.

 

  Fo= r bidirectional chains, is NLF chronology always reversed for each directio= n, or is it arbitrarily  controlled?

[Med] The current text allows to specify distinct NLFC = Map for each direction.

 

Are = NLF’s identified by IP address, FQDN, or both?

<= p class=3DMsoNormal>[Med] I guess your question is about locator= s not identifiers. The current text says both can be supported but we recom= mend the use of an IP address since no resolution capability would be neede= d in the NLF-enabled node. Resolution capabilities (and advanced TE functio= ns) can be enabled only at the PDP (Policy Decision Point); no need to over= load NLF-enabled nodes with such features.

 

Is it an objective for NLFC to explicitly handle load balanci= ng towards multiply-instantiated NLF’s , or is load balancing in this= regard considered to be part of each NLF?

[Med] Yes. The current text says:

 

=A0=A0 o=A0 Advanced features (e.g., load balancing objectives) are als= o

=A0=A0=A0=A0=A0 policy-based.=A0 I= nvoked enforcement points are not aware of such

=A0=A0=A0=A0=A0 objectives.

[Med] Good point= . FYI, we are currently working on a OSPF extension for this purpose. We wi= ll share the text soon.

 

1.4 Assumptions

All NLFs are reachable via standard IP routing techniques (i.e., no as= sumptions regarding non-IP SDN protocols or techniques)?<= /p>

[Med] How FIBs and RIBs are populated is= orthogonal to the overall procedure.

 

1.5 Rationale

Suggest comment about fragmentation be moved to Objectives= section.

[Med] I will= see how to update the text.

 

1.6 Generic Requirements

Whether path stretch occurs or not is an administrative= matter, assuming reasonably correct IP routing topologies.

[Med] Agree. This is deployment-speci= fic.

&nbs= p;

3.3 Building NLF= Maps

Does an = NLF map specify the set of NLF’s, only, or does it specify the set of= NLF’s and the policy to be applied by each.   For example,= assume that Firewall-A and NAT64-A are distinct instances of NLFs. &n= bsp; Is {Firewall-A-profile-1, NAT64-A-profile-1} distinct from {Firewall-A= -profile-2, Nat64-A-profile-1}.   Or, alternatively, is the polic= y conveyed outside of the map?

[Med] I’m not sure I get the point. What do you mean by R= 20;profile” here?

 

4.1  NLFC Boundary Node

<= span lang=3DEN-US>I think the definition of ingress/egress here is problema= tic as it seems to be implying topological direction (i.e., uplink and down= link in a mobile environment).   If, however, the NLFC domain is = thought of as residing in between 2 non-NLFC domains (or segments of the sa= me domain), then there is both ingress and egress at both ends of the chain= .   A more descriptive approach may be to use terms such as ̶= 0;West-ingress”, “East-egress”, “East-ingress”= ;, “West-egress”.    In a wireless domain, perha= ps “north” and “south” would be more intuitive, but= I think the connotation is more limiting in a general sense.

[Med] Each NLFC Boundary Node behav= es in both the ingress and egress roles. It behaves as “ingress”= ; for traffic entering the domain and as “egress” for traffic e= xisting the domain.

 

4= .3 NLFC Ingress Node

As an ingress node which connects to the non NLFC domain, when wou= ld an NLFC Map Index be present?   Also, this implies an encapsul= ation technique that has not been previously introduced in the document.

[Med] Just like setting/= re-setting DSCP values at the boundaries of a domain, the NLFC Map Index ca= n be set by any “non-legitimate” node. This is even a basic att= ack to disturb a service or try granting access to non-authorized services.= This is why we recommend to enforce a basic policy to not trust any NLFC M= ap Index received from adjacent domains.

 

5.1.1 NLFC Map Index

If the individual entries in an NLFC Map indica= te only the identity of the NLF, then 8 or 16 bits to identify which NLFC M= ap would be sufficient.   If, however, the entries in an NLFC Map= indicate NLF identity+policy, then 16 or 32 bits would be more appropriate= .

[Med] The policy is = not signaled in packets; only the index of the NLFC Map is. =A0<= /span>

 <= /p>

5.1.3 Where to store NLFC Map In= dex in a packet?

Consider usage of GRE key.   Advantage is that in many cases, = encapsulation is required and GRE is a standard encapsulation technique whi= ch provides a 32-bit general purpose field.   Disadvantage is tha= t some systems may require each unique GRE key to be provisioned as an IP i= nterface.

[Med] Will a= dd a new entry for it. Another inconvenient I see is for cases where encaps= ulation is not needed at all. This is typically when the next NLF node is a= n adjacent node (see Section 6.2).

 

= --_000_94C682931C08B048B7A8645303FDC9F36EDB5C7F8FPUEXCB1Bnante_-- From christian.jacquenet@orange.com Tue Jun 25 23:39:04 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6BFB221E8119 for ; Tue, 25 Jun 2013 23:39:04 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.247 X-Spam-Level: X-Spam-Status: No, score=-2.247 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HELO_EQ_FR=0.35, HTML_MESSAGE=0.001, UNPARSEABLE_RELAY=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id YZOHQIiNRK3Q for ; Tue, 25 Jun 2013 23:38:59 -0700 (PDT) Received: from relais-inet.francetelecom.com (relais-ias92.francetelecom.com [193.251.215.92]) by ietfa.amsl.com (Postfix) with ESMTP id B702311E815A for ; Tue, 25 Jun 2013 23:38:58 -0700 (PDT) Received: from omfedm05.si.francetelecom.fr (unknown [xx.xx.xx.1]) by omfedm10.si.francetelecom.fr (ESMTP service) with ESMTP id 930A3264AD2; Wed, 26 Jun 2013 08:38:57 +0200 (CEST) Received: from puexch31.nanterre.francetelecom.fr (unknown [10.101.44.29]) by omfedm05.si.francetelecom.fr (ESMTP service) with ESMTP id 75EC535C058; Wed, 26 Jun 2013 08:38:57 +0200 (CEST) Received: from PUEXCB1C.nanterre.francetelecom.fr ([10.101.44.7]) by puexch31.nanterre.francetelecom.fr ([10.101.44.29]) with mapi; Wed, 26 Jun 2013 08:38:57 +0200 From: To: Ron Parker , "nsc@ietf.org" Date: Wed, 26 Jun 2013 08:38:55 +0200 Thread-Topic: Comments on draft-boucadair-network-function-chaining-00 Thread-Index: Ac5x6/wxcUX+ReJaT8eNlB4kpr2L6QASIKPA Message-ID: <32603_1372228737_51CA8C81_32603_2651_1_983A1D8DA0DA5F4EB747BF34CBEE5CD15C418F0433@PUEXCB1C.nanterre.francetelecom.fr> References: In-Reply-To: Accept-Language: fr-FR Content-Language: fr-FR X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: fr-FR Content-Type: multipart/alternative; boundary="_000_983A1D8DA0DA5F4EB747BF34CBEE5CD15C418F0433PUEXCB1Cnante_" MIME-Version: 1.0 X-PMX-Version: 5.6.1.2065439, Antispam-Engine: 2.7.2.376379, Antispam-Data: 2013.6.26.61516 Subject: Re: [nsc] Comments on draft-boucadair-network-function-chaining-00 X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 06:39:04 -0000 --_000_983A1D8DA0DA5F4EB747BF34CBEE5CD15C418F0433PUEXCB1Cnante_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Hello Ron, Many thanks for your feedback - please see inline. [snip] 1.3 Objectives Is it desirable to support unidirectional chains or are only bidirectional = chains considered? [CJ] I guess this depends on the nature of the service to be delivered and = the corresponding NLF functions to be invoked. Admittedly, the draft was or= iginally written with a kind of "request/response" type of service in mind,= but I think nothing really prevents unidirectional chaining scenarios - th= is could apply to a multicast service within the context of a QoS-based mul= ticast forwarding scheme that would solicit several functions in the access= infrastructure to make sure the multicast contents are properly delivered = to the receiver, for example. For bidirectional chains, is NLF chronology always reversed for each dir= ection, or is it arbitrarily controlled? [CJ] This should be arbitrarily controlled. Are NLF's identified by IP address, FQDN, or both? [CJ] I guess you refer to the NKF Identifier, as introduced in section 2, r= ight? As for the NLF Locator, we currently privilege the IP address, but FQ= DN could be used (see Section 3.2). As for the Identifier, this could be bo= th an IP address and a FQDN. Is it an objective for NLFC to explicitly handle load balancing towards mul= tiply-instantiated NLF's , or is load balancing in this regard considered t= o be part of each NLF? [CJ] I think it's primarily a design choice that will be conditioned by the= very nature of the NLF to be activated. I'm not sure each and every NLF we= can think about has implicit load balancing capabilities. Therefore, I wou= ld consider load balancing as a possible (but not necessarily mandatory) ob= jective of the NLFC machinery. Is it an objective for NLFC to determine the availability of an NLF? [CJ] No, I think the availability of an NLF suggests another mechanism, e.g= ., the (automated) discovery of the network topology, devices and their res= pective capabilities which will feed the PDP intelligence with information = to not only dynamically structure a network service but also to trigger NLF= C according to the outcomes of the negotiation of the service parameters be= tween the customer and the provider, for example. If so, are there policies regarding what to do when an NLF becomes unava= ilable? [CJ] To me, this is part of a feedback mechanism introduced in draft-sin-sd= nrg* which is meant to assess how efficiently a policy (or a set thereof) i= s enforced, but also to notify the PDP about any event (such as the unavail= ability of a NLF which failed the chaining operation for example) that may = affect the decision-making process. 1.4 Assumptions All NLFs are reachable via standard IP routing techniques (i.e., no assumpt= ions regarding non-IP SDN protocols or techniques)? [CJ] Yes, the draft is generic enough to consider any IP networking environ= ment. 1.5 Rationale Suggest comment about fragmentation be moved to Objectives section. [CJ] Makes sense. 1.6 Generic Requirements Whether path stretch occurs or not is an administrative matter, assuming re= asonably correct IP routing topologies. [CJ] Agreed, but do you suggest any particular wording? 3.3 Building NLF Maps Does an NLF map specify the set of NLF's, only, or does it specify the set = of NLF's and the policy to be applied by each. For example, assume that F= irewall-A and NAT64-A are distinct instances of NLFs. Is {Firewall-A-prof= ile-1, NAT64-A-profile-1} distinct from {Firewall-A-profile-2, Nat64-A-prof= ile-1}. Or, alternatively, is the policy conveyed outside of the map? [CJ] Policies are conveyed outside of the map. 4.1 NLFC Boundary Node I think the definition of ingress/egress here is problematic as it seems to= be implying topological direction (i.e., uplink and downlink in a mobile e= nvironment). If, however, the NLFC domain is thought of as residing in be= tween 2 non-NLFC domains (or segments of the same domain), then there is bo= th ingress and egress at both ends of the chain. A more descriptive appro= ach may be to use terms such as "West-ingress", "East-egress", "East-ingres= s", "West-egress". In a wireless domain, perhaps "north" and "south" wou= ld be more intuitive, but I think the connotation is more limiting in a gen= eral sense. [CJ] I would avoid that kind of elaboration. As for the case you rightfully= mention, I think we can still deal with ingress and egress notions on both= "sides" of the NLFC domain. 4.3 NLFC Ingress Node As an ingress node which connects to the non NLFC domain, when would an NLF= C Map Index be present? [CJ] I guess the operation is similar to DSCP-re-marking, with implicitly s= ecurity threats that have been mentioned in the document. Also, this implies an encapsulation technique that has not been previousl= y introduced in the document. 5.1.1 NLFC Map Index If the individual entries in an NLFC Map indicate only the identity of the = NLF, then 8 or 16 bits to identify which NLFC Map would be sufficient. If= , however, the entries in an NLFC Map indicate NLF identity+policy, then 16= or 32 bits would be more appropriate. [CJ] Right. 5.1.3 Where to store NLFC Map Index in a packet? Consider usage of GRE key. Advantage is that in many cases, encapsulation= is required and GRE is a standard encapsulation technique which provides a= 32-bit general purpose field. Disadvantage is that some systems may requ= ire each unique GRE key to be provisioned as an IP interface. [CJ] OK. Thanks, Christian. ___________________________________________________________________________= ______________________________________________ Ce message et ses pieces jointes peuvent contenir des informations confiden= tielles ou privilegiees et ne doivent donc pas etre diffuses, exploites ou copies sans autorisation. Si vous avez recu= ce message par erreur, veuillez le signaler a l'expediteur et le detruire ainsi que les pieces jointes. Les messages el= ectroniques etant susceptibles d'alteration, Orange decline toute responsabilite si ce message a ete altere, deforme ou = falsifie. Merci. This message and its attachments may contain confidential or privileged inf= ormation that may be protected by law; they should not be distributed, used or copied without authorisation. If you have received this email in error, please notify the sender and dele= te this message and its attachments. As emails may be altered, Orange is not liable for messages that have been = modified, changed or falsified. Thank you. --_000_983A1D8DA0DA5F4EB747BF34CBEE5CD15C418F0433PUEXCB1Cnante_ Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable

Hello Ron,=

 

Many thanks for your feedback – please see inline.<= /o:p>

 

[snip]<= o:p>

 =

1.3 Objectives<= o:p>

Is it desir= able to support unidirectional chains or are only bidirectional chains cons= idered?

[CJ] I guess this depends on the nature of the service to be delivered a= nd the corresponding NLF functions to be invoked. Admittedly, the draft was= originally written with a kind of “request/response” type of s= ervice in mind, but I think nothing really prevents unidirectional chaining= scenarios – this could apply to a multicast service within the conte= xt of a QoS-based multicast forwarding scheme that would solicit several fu= nctions in the access infrastructure to make sure the multicast contents ar= e properly delivered to the receiver, for example.

   For bidirection= al chains, is NLF chronology always reversed for each direction, or is it a= rbitrarily  controlled?

[CJ] This sho= uld be arbitrarily controlled.

Are NLF’s identified by IP address, FQDN, = or both?

[CJ] I guess you refer to the NKF= Identifier, as introduced in section 2, right? As for the NLF Locator, we = currently privilege the IP address, but FQDN could be used (see Section 3.2= ). As for the Identifier, this could be both an IP address and a FQDN.=

Is it an= objective for NLFC to explicitly handle load balancing towards multiply-in= stantiated NLF’s , or is load balancing in this regard considered to = be part of each NLF?

[CJ] I think it’= ;s primarily a design choice that will be conditioned by the very nature of= the NLF to be activated. I’m not sure each and every NLF we can thin= k about has implicit load balancing capabilities. Therefore, I would consid= er load balancing as a possible (but not necessarily mandatory) objective o= f the NLFC machinery.

Is it an objective for NLFC to determine the availability= of an NLF?

[CJ] No, I think the availability of an NLF suggests another mechani= sm, e.g., the (automated) discovery of the network topology, devices and th= eir respective capabilities which will feed the PDP intelligence with infor= mation to not only dynamically structure a network service but also to trig= ger NLFC according to the outcomes of the negotiation of the service parame= ters between the customer and the provider, for example.<= /p>

   If so, ar= e there policies regarding what to do when an NLF becomes unavailable?=

[CJ] To me, this is part of a feedback mechani= sm introduced in draft-sin-sdnrg* which is meant to assess how efficiently = a policy (or a set thereof) is enforced, but also to notify the PDP about a= ny event (such as the unavailability of a NLF which failed the chaining ope= ration for example) that may affect the decision-making process.=

 

1.4 Assumptions

All NLFs are = reachable via standard IP routing techniques (i.e., no assumptions regardin= g non-IP SDN protocols or techniques)?

<= span style=3D'font-size:10.0pt;font-family:"Courier New";color:#7030A0'>[CJ= ] Yes, the draft is generic enough to consider any IP networking environmen= t.

<= o:p> 

1.5 R= ationale

Su= ggest comment about fragmentation be moved to Objectives section.

[CJ] Makes sense.

 

1.6 Generic Requirements

Whether path stretch occurs= or not is an administrative matter, assuming reasonably correct IP routing= topologies.

[CJ] Agreed, but do you sugge= st any particular wording?

 

3.3 Building NLF Maps

Does an NLF map specify the set of NLF’= ;s, only, or does it specify the set of NLF’s and the policy to be ap= plied by each.   For example, assume that Firewall-A and NAT64-A = are distinct instances of NLFs.   Is {Firewall-A-profile-1, NAT64= -A-profile-1} distinct from {Firewall-A-profile-2, Nat64-A-profile-1}. = ;  Or, alternatively, is the policy conveyed outside of the map?<= /o:p>

[CJ] Policies are conveyed outside of the map.<= o:p>

 

4.1 = ; NLFC Boundary Node

I think the definition of ingress/egress here is problematic as i= t seems to be implying topological direction (i.e., uplink and downlink in = a mobile environment).   If, however, the NLFC domain is thought = of as residing in between 2 non-NLFC domains (or segments of the same domai= n), then there is both ingress and egress at both ends of the chain. &= nbsp; A more descriptive approach may be to use terms such as “West-i= ngress”, “East-egress”, “East-ingress”, ̶= 0;West-egress”.    In a wireless domain, perhaps ̶= 0;north” and “south” would be more intuitive, but I think= the connotation is more limiting in a general sense.

[CJ] I would avoid that kind of elaboration. As for the case yo= u rightfully mention, I think we can still deal with ingress and egress not= ions on both “sides” of the NLFC domain.

<= p class=3DMsoNormal style=3D'margin-left:36.0pt'> 

4.3 NLFC Ingress Node

As an ingress node = which connects to the non NLFC domain, when would an NLFC Map Index be pres= ent? 

[CJ] I guess the operation is similar to DSCP-re-marking, with implic= itly security threats that have been mentioned in the document.<= /span>

  Also, th= is implies an encapsulation technique that has not been previously introduc= ed in the document.

 

5.1.1 NLFC Map Index

If the individual entries in an NLFC Map indicate only the i= dentity of the NLF, then 8 or 16 bits to identify which NLFC Map would be s= ufficient.   If, however, the entries in an NLFC Map indicate NLF= identity+policy, then 16 or 32 bits would be more appropriate.<= /p>

[CJ] Right.

 

5.1.3 Where to store NLFC Map Index in a packet?<= o:p>

Consider us= age of GRE key.   Advantage is that in many cases, encapsulation = is required and GRE is a standard encapsulation technique which provides a = 32-bit general purpose field.   Disadvantage is that some systems= may require each unique GRE key to be provisioned as an IP interface.=

[CJ] OK.

 

Thanks,=

 

Chris= tian.

 

_________________________________________=
___________________________________________________________________________=
_____

Ce message et ses pieces jointes peuvent contenir des informations confiden=
tielles ou privilegiees et ne doivent donc
pas etre diffuses, exploites ou copies sans autorisation. Si vous avez recu=
 ce message par erreur, veuillez le signaler
a l'expediteur et le detruire ainsi que les pieces jointes. Les messages el=
ectroniques etant susceptibles d'alteration,
Orange decline toute responsabilite si ce message a ete altere, deforme ou =
falsifie. Merci.

This message and its attachments may contain confidential or privileged inf=
ormation that may be protected by law;
they should not be distributed, used or copied without authorisation.
If you have received this email in error, please notify the sender and dele=
te this message and its attachments.
As emails may be altered, Orange is not liable for messages that have been =
modified, changed or falsified.
Thank you.
= --_000_983A1D8DA0DA5F4EB747BF34CBEE5CD15C418F0433PUEXCB1Cnante_-- From jguichar@cisco.com Wed Jun 26 05:26:03 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id AFA0321F8267 for ; Wed, 26 Jun 2013 05:25:58 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.599 X-Spam-Level: X-Spam-Status: No, score=-10.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id yEE7xX3w6B+X for ; Wed, 26 Jun 2013 05:25:54 -0700 (PDT) Received: from rcdn-iport-4.cisco.com (rcdn-iport-4.cisco.com [173.37.86.75]) by ietfa.amsl.com (Postfix) with ESMTP id 76B5421F85B2 for ; Wed, 26 Jun 2013 05:25:50 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=7265; q=dns/txt; s=iport; t=1372249550; x=1373459150; h=from:to:cc:subject:date:message-id:in-reply-to: content-id:content-transfer-encoding:mime-version; bh=jrI67/Wr/q1dDwoKRhCZomdViavgzPAASfYt2q4Wqzw=; b=g2N0sPQ/GR5ld0K11RJaoinfOqhLazkDIbqKQ4T7w8Cmh5HMLQ8XW4zn ms9m88n+a+zdMusmCSi8uJQvkSjBxee9cN1qKG8XWkalLGgunnM7OFQAH caGEkcwEKiIQaY0QuVg8HbFcGlQLiSWlZVVwM+A3knBtODT5Bp6fWWTxt g=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AhIFAPbcylGtJXG+/2dsb2JhbABagwkxSb1xgQqBAxZ0giMBAQEDAQEBAWsLBQcGAQgRBAEBAQodLgsUCQgCBAENBQiIAAYMukCOCRB/AjEHBoJ8YQOIapAEiQaHFoMRgXE3 X-IronPort-AV: E=Sophos;i="4.87,943,1363132800"; d="scan'208";a="227699115" Received: from rcdn-core2-3.cisco.com ([173.37.113.190]) by rcdn-iport-4.cisco.com with ESMTP; 26 Jun 2013 12:25:49 +0000 Received: from xhc-rcd-x15.cisco.com (xhc-rcd-x15.cisco.com [173.37.183.89]) by rcdn-core2-3.cisco.com (8.14.5/8.14.5) with ESMTP id r5QCPno8026055 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Wed, 26 Jun 2013 12:25:49 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.80]) by xhc-rcd-x15.cisco.com ([173.37.183.89]) with mapi id 14.02.0318.004; Wed, 26 Jun 2013 07:25:49 -0500 From: "Jim Guichard (jguichar)" To: "Hongyu Li (Julio)" , "Diego R. Lopez" , " " Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewABC13ID//+4ggP//KpNAgAYtNICAA2WvAIAAXX0A Date: Wed, 26 Jun 2013 12:25:48 +0000 Message-ID: <68B171751455884590F8E38E96416F363F556D43@xmb-rcd-x01.cisco.com> In-Reply-To: <6EB34CB5D82C4645B826C56144826EA97DF0AD29@szxeml560-mbx.china.huawei.com> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.2.3.120616 x-originating-ip: [10.98.130.212] Content-Type: text/plain; charset="iso-8859-1" Content-ID: <14418415B20F3E429ED779294FB08351@emea.cisco.com> Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Cc: "nsc@ietf.org" , "adrian@olddog.co.uk" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 12:26:03 -0000 Hi Hongyu, Comments inline .. On 6/25/13 10:51 PM, "Hongyu Li (Julio)" wrote: >Hi, > >Please allow me to chip in on the original questions. > >NSC may be interpreted in two ways. One is what's chained are network >functions.=20 Jim> I would say that what is chained are functions applied to network traffic.=20 >The other is service functions are chained on network layer. Both should >be in the scope. In other words, we are seeking chaining SFs on network >level, while SF could be network service functions or higher layer >service functions. Jim> yes, exactly right. Perhaps the correct terminology is Lx-L7 service functions? =20 > >Another point I wanna make is that IMHO, on key character of NSC is >traffic is forwarded per service characteristics other than destination >addresses. Jim> it is certainly possible to form basic service chains using only destination address through manipulation of routing. However, more complex chains that reflect service characteristics and policy, or service functions that are able to consume and/or insert metadata require a more flexible steering mechanism as you say. > This is the key value NSC WG, if will be formed, should work on. Service >characteristics may include the network service a use has subscribed to, >e.g. parental control, firewall. That means a packet is forwarded into a >service chain, when a classifier matches some rules of service >characteristics, which may be reflected by access ports, or segments or >tuples in a packet. > >Cheers, >Hongyu > >-----Original Message----- >From: nsc-bounces@ietf.org [mailto:nsc-bounces@ietf.org] On Behalf Of >Diego R. Lopez >Sent: Monday, June 24, 2013 6:58 AM >To: >Cc: nsc@ietf.org; Jim Guichard (jguichar); adrian@olddog.co.uk >Subject: Re: [nsc] What is a service? Why is NSC not SC? > >Hi all, > >Looking at this thread, combining the distinction between "function" and >"service" provided by Benson, and the assumptions Med lists in what >relates to how policy applies to them within an administrative domain, I >am in favor of talking of functions that are combined (chained) to build >services, at any recursion level: a service can be composed (chained) as >a function to build a more complex service. > >Probably the Solomonic proposal from Jim is the one making more sense... > >Be goode, > >On 21 Jun 2013, at 09:45 , > wrote: > >> Hi Jim, all, >> >> I would even say: these functions are seen as black-boxes and no >>assumption is to be made on the outcome of such functions. >> >> We tried to capture this and other assumptions in draft-boucadair-* >> >> o NLFs are handled as black-boxes; the technical characterization of >> each NLF is not required. >> o Not all NLFs can be characterized with a standard definition in >> terms of technical description, detailed specification, etc. >> o There is no global nor standard list of NLFs enabled in a given >> administrative domain. The set of NLFs varies on a per deployment >> basis. >> o There is no global nor standard NLF chaining logic. The ordered >> set of NLFs that need to be activated to deliver a given >> connectivity service is specific to each administrative entity. >> o The chaining of NLFs and the criteria to invoke some of them are >> local to each administrative entity that operates a connectivity >> network (also called administrative domain). >> o NLF chaining logic and related policies should not be exposed >> outside a given administrative domain. >> o Several NLF chaining logics can be simultaneously enforced within >> an administrative domain to meet business requirements. >> >> Cheers, >> Med >> >>> -----Message d'origine----- >>> De : nsc-bounces@ietf.org [mailto:nsc-bounces@ietf.org] De la part de >>>Jim >>> Guichard (jguichar) >>> Envoy=E9 : jeudi 20 juin 2013 19:23 >>> =C0 : adrian@olddog.co.uk; nsc@ietf.org >>> Objet : Re: [nsc] What is a service? Why is NSC not SC? >>> >>> Hi Adrian, >>> >>> On 6/20/13 10:26 AM, "Adrian Farrel" wrote: >>> >>>> Thanks to Ken and Med for replies. >>>> >>>> Choosing to respond on Med's thread because he gave... >>>> >>>>> o NLF (Network-Located Function): refers to a function which is >>>>> enabled in the network operated by an administrative entity. >>>>>NLF >>>>> can be for example: firewall (e.g., [RFC6092]), DPI (Deep Packet >>>>> Inspection), LI (Lawful Intercept) module, NAT44 [RFC3022], >>>>>NAT64 >>>>> [RFC6146], DS-Lite AFTR [RFC6333], NPTv6 [RFC6296], HOST_ID >>>>> injection, HTTP Header Enrichment function, TCP optimizer, etc. >>>> >>>> Something I pick out of this is that all the example functions are >>>> packet-oriented functions (or possibly flow-oriented) and not >>>>application >>>> layer >>>> services. I think this is significant and not accidental :-) >>> >>> Jim> actually the intent was that both packet-oriented and application >>> layer service functions are within scope. It is certainly possible to >>> envisage service chains that incorporate both of these types of >>>functions >>> and we should not restrict our work to one or the other but rather let >>>the >>> use cases dictate. >>> >>> >>>> Can we capture that idea in a definition? >>>> >>>> I note that it is hinted at in Paul and Jim's BoF proposal text and >>>>it is >>>> there >>>> because I asked them why this proposal was aimed at the Routing Area >>>>not >>>> the >>>> Applications Area. To which the answer was "because we want to steer >>>> packets or >>>> flows to the place where the next service will be applied." >>>> >>>> Adrian >>>> >>>> _______________________________________________ >>>> nsc mailing list >>>> nsc@ietf.org >>>> https://www.ietf.org/mailman/listinfo/nsc >>> >>> _______________________________________________ >>> nsc mailing list >>> nsc@ietf.org >>> https://www.ietf.org/mailman/listinfo/nsc >> _______________________________________________ >> nsc mailing list >> nsc@ietf.org >> https://www.ietf.org/mailman/listinfo/nsc > > >-- >"Esta vez no fallaremos, Doctor Infierno" > >Dr Diego R. Lopez >Telefonica I+D >http://people.tid.es/diego.lopez/ > >e-mail: diego@tid.es >Tel: +34 913 129 041 >Mobile: +34 682 051 091 >----------------------------------------- > > >________________________________ > >Este mensaje se dirige exclusivamente a su destinatario. Puede consultar >nuestra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el = enlace >situado m=E1s abajo. >This message is intended exclusively for its addressee. We only send and >receive email on the basis of the terms set out at: >http://www.tid.es/ES/PAGINAS/disclaimer.aspx >_______________________________________________ >nsc mailing list >nsc@ietf.org >https://www.ietf.org/mailman/listinfo/nsc >_______________________________________________ >nsc mailing list >nsc@ietf.org >https://www.ietf.org/mailman/listinfo/nsc From Ron_Parker@affirmednetworks.com Wed Jun 26 06:13:44 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 3242C21E8122 for ; Wed, 26 Jun 2013 06:13:44 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.599 X-Spam-Level: X-Spam-Status: No, score=-2.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id toX0YBzy3LGE for ; Wed, 26 Jun 2013 06:13:39 -0700 (PDT) Received: from hub021-ca-6.exch021.serverdata.net (hub021-ca-6.exch021.serverdata.net [64.78.56.71]) by ietfa.amsl.com (Postfix) with ESMTP id B5DF621E810B for ; Wed, 26 Jun 2013 06:13:39 -0700 (PDT) Received: from MBX021-W3-CA-2.exch021.domain.local ([10.254.4.78]) by HUB021-CA-6.exch021.domain.local ([10.254.4.92]) with mapi id 14.03.0123.003; Wed, 26 Jun 2013 06:13:37 -0700 From: Ron Parker To: "Hongyu Li (Julio)" , "Diego R. Lopez" , " " Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrew//+rwICAADExAIAA8PYAgAQjwYD//Bu+0P/3g9Mg Date: Wed, 26 Jun 2013 13:13:37 +0000 Message-ID: References: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> <6EB34CB5D82C4645B826C56144826EA97DF0AD29@szxeml560-mbx.china.huawei.com> In-Reply-To: <6EB34CB5D82C4645B826C56144826EA97DF0AD29@szxeml560-mbx.china.huawei.com> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [50.203.66.100] Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: base64 MIME-Version: 1.0 Cc: "nsc@ietf.org" , "Jim Guichard \(jguichar\)" , "adrian@olddog.co.uk" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 13:13:44 -0000 SG9uZ3l1LCByZWdhcmRpbmc6DQoNCltzbmlwXSBBbm90aGVyIHBvaW50IEkgd2FubmEgbWFrZSBp cyB0aGF0IElNSE8sIG9uIGtleSBjaGFyYWN0ZXIgb2YgTlNDIGlzIHRyYWZmaWMgaXMgZm9yd2Fy ZGVkIHBlciBzZXJ2aWNlIGNoYXJhY3RlcmlzdGljcyBvdGhlciB0aGFuIGRlc3RpbmF0aW9uIGFk ZHJlc3Nlcy4gVGhpcyBpcyB0aGUga2V5IHZhbHVlIE5TQyBXRywgaWYgd2lsbCBiZSBmb3JtZWQs IHNob3VsZCB3b3JrIG9uLiBTZXJ2aWNlIGNoYXJhY3RlcmlzdGljcyBtYXkgaW5jbHVkZSB0aGUg bmV0d29yayBzZXJ2aWNlIGEgdXNlIGhhcyBzdWJzY3JpYmVkIHRvLCBlLmcuIHBhcmVudGFsIGNv bnRyb2wsIGZpcmV3YWxsLiBUaGF0IG1lYW5zIGEgcGFja2V0IGlzIGZvcndhcmRlZCBpbnRvIGEg c2VydmljZSBjaGFpbiwgd2hlbiBhIGNsYXNzaWZpZXIgbWF0Y2hlcyBzb21lIHJ1bGVzIG9mIHNl cnZpY2UgY2hhcmFjdGVyaXN0aWNzLCB3aGljaCBtYXkgYmUgcmVmbGVjdGVkIGJ5IGFjY2VzcyBw b3J0cywgb3Igc2VnbWVudHMgb3IgdHVwbGVzIGluIGEgcGFja2V0Lg0KDQpbUm9uXTogIEkgdmll dyBOU0MgYXMgYSBzcGVjaWZpYyBhcHBsaWNhdGlvbiBvZiBwb2xpY3kgYmFzZWQgZm9yd2FyZGlu Zywgd2hlcmUgdGhlIHBvbGljeSBpcyBiYXNlZCBvbiBhbnkgb3IgYWxsIG9mIHRoZSBmaWVsZHMg aW4gdGhlIElQIGhlYWRlci4gICBUaGUgZW5kLXRvLWVuZCBtb2RlbCBvZiBjbGFzc2lmaWVyLWJh c2VkIG5vZGVzIGFuZCBub24tY2xhc3NpZmllci1iYXNlZCBub2RlcyBtZWFucyB0aGF0IHRoZXJl IGFyZSAyIGRpc3RpbmN0IHNldHMgb2YgcHJvY2VkdXJlcyB0byBkZXRlcm1pbmUgYW5kIGFwcGx5 IHRoZSBwb2xpY3kuDQoNCg0KQmVzdCByZWdhcmRzLA0KDQpSb24NCg0KDQotLS0tLU9yaWdpbmFs IE1lc3NhZ2UtLS0tLQ0KRnJvbTogbnNjLWJvdW5jZXNAaWV0Zi5vcmcgW21haWx0bzpuc2MtYm91 bmNlc0BpZXRmLm9yZ10gT24gQmVoYWxmIE9mIEhvbmd5dSBMaSAoSnVsaW8pDQpTZW50OiBUdWVz ZGF5LCBKdW5lIDI1LCAyMDEzIDEwOjUxIFBNDQpUbzogRGllZ28gUi4gTG9wZXo7IDxtb2hhbWVk LmJvdWNhZGFpckBvcmFuZ2UuY29tPiA8bW9oYW1lZC5ib3VjYWRhaXJAb3JhbmdlLmNvbT4NCkNj OiBuc2NAaWV0Zi5vcmc7IEppbSBHdWljaGFyZCAoamd1aWNoYXIpOyBhZHJpYW5Ab2xkZG9nLmNv LnVrDQpTdWJqZWN0OiBSZTogW25zY10gV2hhdCBpcyBhIHNlcnZpY2U/IFdoeSBpcyBOU0Mgbm90 IFNDPw0KDQpIaSwNCg0KUGxlYXNlIGFsbG93IG1lIHRvIGNoaXAgaW4gb24gdGhlIG9yaWdpbmFs IHF1ZXN0aW9ucy4NCg0KTlNDIG1heSBiZSBpbnRlcnByZXRlZCBpbiB0d28gd2F5cy4gT25lIGlz IHdoYXQncyBjaGFpbmVkIGFyZSBuZXR3b3JrIGZ1bmN0aW9ucy4gVGhlIG90aGVyIGlzIHNlcnZp Y2UgZnVuY3Rpb25zIGFyZSBjaGFpbmVkIG9uIG5ldHdvcmsgbGF5ZXIuIEJvdGggc2hvdWxkIGJl IGluIHRoZSBzY29wZS4gSW4gb3RoZXIgd29yZHMsIHdlIGFyZSBzZWVraW5nIGNoYWluaW5nIFNG cyBvbiBuZXR3b3JrIGxldmVsLCB3aGlsZSBTRiBjb3VsZCBiZSBuZXR3b3JrIHNlcnZpY2UgZnVu Y3Rpb25zIG9yIGhpZ2hlciBsYXllciBzZXJ2aWNlIGZ1bmN0aW9ucy4NCg0KQW5vdGhlciBwb2lu dCBJIHdhbm5hIG1ha2UgaXMgdGhhdCBJTUhPLCBvbiBrZXkgY2hhcmFjdGVyIG9mIE5TQyBpcyB0 cmFmZmljIGlzIGZvcndhcmRlZCBwZXIgc2VydmljZSBjaGFyYWN0ZXJpc3RpY3Mgb3RoZXIgdGhh biBkZXN0aW5hdGlvbiBhZGRyZXNzZXMuIFRoaXMgaXMgdGhlIGtleSB2YWx1ZSBOU0MgV0csIGlm IHdpbGwgYmUgZm9ybWVkLCBzaG91bGQgd29yayBvbi4gU2VydmljZSBjaGFyYWN0ZXJpc3RpY3Mg bWF5IGluY2x1ZGUgdGhlIG5ldHdvcmsgc2VydmljZSBhIHVzZSBoYXMgc3Vic2NyaWJlZCB0bywg ZS5nLiBwYXJlbnRhbCBjb250cm9sLCBmaXJld2FsbC4gVGhhdCBtZWFucyBhIHBhY2tldCBpcyBm b3J3YXJkZWQgaW50byBhIHNlcnZpY2UgY2hhaW4sIHdoZW4gYSBjbGFzc2lmaWVyIG1hdGNoZXMg c29tZSBydWxlcyBvZiBzZXJ2aWNlIGNoYXJhY3RlcmlzdGljcywgd2hpY2ggbWF5IGJlIHJlZmxl Y3RlZCBieSBhY2Nlc3MgcG9ydHMsIG9yIHNlZ21lbnRzIG9yIHR1cGxlcyBpbiBhIHBhY2tldC4N Cg0KQ2hlZXJzLA0KSG9uZ3l1DQoNCi0tLS0tT3JpZ2luYWwgTWVzc2FnZS0tLS0tDQpGcm9tOiBu c2MtYm91bmNlc0BpZXRmLm9yZyBbbWFpbHRvOm5zYy1ib3VuY2VzQGlldGYub3JnXSBPbiBCZWhh bGYgT2YgRGllZ28gUi4gTG9wZXoNClNlbnQ6IE1vbmRheSwgSnVuZSAyNCwgMjAxMyA2OjU4IEFN DQpUbzogPG1vaGFtZWQuYm91Y2FkYWlyQG9yYW5nZS5jb20+IDxtb2hhbWVkLmJvdWNhZGFpckBv cmFuZ2UuY29tPg0KQ2M6IG5zY0BpZXRmLm9yZzsgSmltIEd1aWNoYXJkIChqZ3VpY2hhcik7IGFk cmlhbkBvbGRkb2cuY28udWsNClN1YmplY3Q6IFJlOiBbbnNjXSBXaGF0IGlzIGEgc2VydmljZT8g V2h5IGlzIE5TQyBub3QgU0M/DQoNCkhpIGFsbCwNCg0KTG9va2luZyBhdCB0aGlzIHRocmVhZCwg Y29tYmluaW5nIHRoZSBkaXN0aW5jdGlvbiBiZXR3ZWVuICJmdW5jdGlvbiIgYW5kICJzZXJ2aWNl IiBwcm92aWRlZCBieSBCZW5zb24sIGFuZCB0aGUgYXNzdW1wdGlvbnMgTWVkIGxpc3RzIGluIHdo YXQgcmVsYXRlcyB0byBob3cgcG9saWN5IGFwcGxpZXMgdG8gdGhlbSB3aXRoaW4gYW4gYWRtaW5p c3RyYXRpdmUgZG9tYWluLCBJIGFtIGluIGZhdm9yIG9mIHRhbGtpbmcgb2YgZnVuY3Rpb25zIHRo YXQgYXJlIGNvbWJpbmVkIChjaGFpbmVkKSB0byBidWlsZCBzZXJ2aWNlcywgYXQgYW55IHJlY3Vy c2lvbiBsZXZlbDogYSBzZXJ2aWNlIGNhbiBiZSBjb21wb3NlZCAoY2hhaW5lZCkgYXMgYSBmdW5j dGlvbiB0byBidWlsZCBhIG1vcmUgY29tcGxleCBzZXJ2aWNlLg0KDQpQcm9iYWJseSB0aGUgU29s b21vbmljIHByb3Bvc2FsIGZyb20gSmltIGlzIHRoZSBvbmUgbWFraW5nIG1vcmUgc2Vuc2UuLi4N Cg0KQmUgZ29vZGUsDQoNCk9uIDIxIEp1biAyMDEzLCBhdCAwOTo0NSAsIDxtb2hhbWVkLmJvdWNh ZGFpckBvcmFuZ2UuY29tPiAgPG1vaGFtZWQuYm91Y2FkYWlyQG9yYW5nZS5jb20+IHdyb3RlOg0K DQo+IEhpIEppbSwgYWxsLA0KPg0KPiBJIHdvdWxkIGV2ZW4gc2F5OiB0aGVzZSBmdW5jdGlvbnMg YXJlIHNlZW4gYXMgYmxhY2stYm94ZXMgYW5kIG5vIGFzc3VtcHRpb24gaXMgdG8gYmUgbWFkZSBv biB0aGUgb3V0Y29tZSBvZiBzdWNoIGZ1bmN0aW9ucy4NCj4NCj4gV2UgdHJpZWQgdG8gY2FwdHVy ZSB0aGlzIGFuZCBvdGhlciBhc3N1bXB0aW9ucyBpbiBkcmFmdC1ib3VjYWRhaXItKg0KPg0KPiAg IG8gIE5MRnMgYXJlIGhhbmRsZWQgYXMgYmxhY2stYm94ZXM7IHRoZSB0ZWNobmljYWwgY2hhcmFj dGVyaXphdGlvbiBvZg0KPiAgICAgIGVhY2ggTkxGIGlzIG5vdCByZXF1aXJlZC4NCj4gICBvICBO b3QgYWxsIE5MRnMgY2FuIGJlIGNoYXJhY3Rlcml6ZWQgd2l0aCBhIHN0YW5kYXJkIGRlZmluaXRp b24gaW4NCj4gICAgICB0ZXJtcyBvZiB0ZWNobmljYWwgZGVzY3JpcHRpb24sIGRldGFpbGVkIHNw ZWNpZmljYXRpb24sIGV0Yy4NCj4gICBvICBUaGVyZSBpcyBubyBnbG9iYWwgbm9yIHN0YW5kYXJk IGxpc3Qgb2YgTkxGcyBlbmFibGVkIGluIGEgZ2l2ZW4NCj4gICAgICBhZG1pbmlzdHJhdGl2ZSBk b21haW4uICBUaGUgc2V0IG9mIE5MRnMgdmFyaWVzIG9uIGEgcGVyIGRlcGxveW1lbnQNCj4gICAg ICBiYXNpcy4NCj4gICBvICBUaGVyZSBpcyBubyBnbG9iYWwgbm9yIHN0YW5kYXJkIE5MRiBjaGFp bmluZyBsb2dpYy4gIFRoZSBvcmRlcmVkDQo+ICAgICAgc2V0IG9mIE5MRnMgdGhhdCBuZWVkIHRv IGJlIGFjdGl2YXRlZCB0byBkZWxpdmVyIGEgZ2l2ZW4NCj4gICAgICBjb25uZWN0aXZpdHkgc2Vy dmljZSBpcyBzcGVjaWZpYyB0byBlYWNoIGFkbWluaXN0cmF0aXZlIGVudGl0eS4NCj4gICBvICBU aGUgY2hhaW5pbmcgb2YgTkxGcyBhbmQgdGhlIGNyaXRlcmlhIHRvIGludm9rZSBzb21lIG9mIHRo ZW0gYXJlDQo+ICAgICAgbG9jYWwgdG8gZWFjaCBhZG1pbmlzdHJhdGl2ZSBlbnRpdHkgdGhhdCBv cGVyYXRlcyBhIGNvbm5lY3Rpdml0eQ0KPiAgICAgIG5ldHdvcmsgKGFsc28gY2FsbGVkIGFkbWlu aXN0cmF0aXZlIGRvbWFpbikuDQo+ICAgbyAgTkxGIGNoYWluaW5nIGxvZ2ljIGFuZCByZWxhdGVk IHBvbGljaWVzIHNob3VsZCBub3QgYmUgZXhwb3NlZA0KPiAgICAgIG91dHNpZGUgYSBnaXZlbiBh ZG1pbmlzdHJhdGl2ZSBkb21haW4uDQo+ICAgbyAgU2V2ZXJhbCBOTEYgY2hhaW5pbmcgbG9naWNz IGNhbiBiZSBzaW11bHRhbmVvdXNseSBlbmZvcmNlZCB3aXRoaW4NCj4gICAgICBhbiBhZG1pbmlz dHJhdGl2ZSBkb21haW4gdG8gbWVldCBidXNpbmVzcyByZXF1aXJlbWVudHMuDQo+DQo+IENoZWVy cywNCj4gTWVkDQo+DQo+PiAtLS0tLU1lc3NhZ2UgZCdvcmlnaW5lLS0tLS0NCj4+IERlIDogbnNj LWJvdW5jZXNAaWV0Zi5vcmcgW21haWx0bzpuc2MtYm91bmNlc0BpZXRmLm9yZ10gRGUgbGEgcGFy dCBkZSANCj4+IEppbSBHdWljaGFyZCAoamd1aWNoYXIpIEVudm95w6kgOiBqZXVkaSAyMCBqdWlu IDIwMTMgMTk6MjMgw4AgOiANCj4+IGFkcmlhbkBvbGRkb2cuY28udWs7IG5zY0BpZXRmLm9yZyBP YmpldCA6IFJlOiBbbnNjXSBXaGF0IGlzIGEgDQo+PiBzZXJ2aWNlPyBXaHkgaXMgTlNDIG5vdCBT Qz8NCj4+DQo+PiBIaSBBZHJpYW4sDQo+Pg0KPj4gT24gNi8yMC8xMyAxMDoyNiBBTSwgIkFkcmlh biBGYXJyZWwiIDxhZHJpYW5Ab2xkZG9nLmNvLnVrPiB3cm90ZToNCj4+DQo+Pj4gVGhhbmtzIHRv IEtlbiBhbmQgTWVkIGZvciByZXBsaWVzLg0KPj4+DQo+Pj4gQ2hvb3NpbmcgdG8gcmVzcG9uZCBv biBNZWQncyB0aHJlYWQgYmVjYXVzZSBoZSBnYXZlLi4uDQo+Pj4NCj4+Pj4gICBvICBOTEYgKE5l dHdvcmstTG9jYXRlZCBGdW5jdGlvbik6IHJlZmVycyB0byBhIGZ1bmN0aW9uIHdoaWNoIGlzDQo+ Pj4+ICAgICAgZW5hYmxlZCBpbiB0aGUgbmV0d29yayBvcGVyYXRlZCBieSBhbiBhZG1pbmlzdHJh dGl2ZSBlbnRpdHkuICBOTEYNCj4+Pj4gICAgICBjYW4gYmUgZm9yIGV4YW1wbGU6IGZpcmV3YWxs IChlLmcuLCBbUkZDNjA5Ml0pLCBEUEkgKERlZXAgUGFja2V0DQo+Pj4+ICAgICAgSW5zcGVjdGlv biksIExJIChMYXdmdWwgSW50ZXJjZXB0KSBtb2R1bGUsIE5BVDQ0IFtSRkMzMDIyXSwgTkFUNjQN Cj4+Pj4gICAgICBbUkZDNjE0Nl0sIERTLUxpdGUgQUZUUiBbUkZDNjMzM10sIE5QVHY2IFtSRkM2 Mjk2XSwgSE9TVF9JRA0KPj4+PiAgICAgIGluamVjdGlvbiwgSFRUUCBIZWFkZXIgRW5yaWNobWVu dCBmdW5jdGlvbiwgVENQIG9wdGltaXplciwgZXRjLg0KPj4+DQo+Pj4gU29tZXRoaW5nIEkgcGlj ayBvdXQgb2YgdGhpcyBpcyB0aGF0IGFsbCB0aGUgZXhhbXBsZSBmdW5jdGlvbnMgYXJlIA0KPj4+ IHBhY2tldC1vcmllbnRlZCBmdW5jdGlvbnMgKG9yIHBvc3NpYmx5IGZsb3ctb3JpZW50ZWQpIGFu ZCBub3QgDQo+Pj4gYXBwbGljYXRpb24gbGF5ZXIgc2VydmljZXMuIEkgdGhpbmsgdGhpcyBpcyBz aWduaWZpY2FudCBhbmQgbm90IA0KPj4+IGFjY2lkZW50YWwgOi0pDQo+Pg0KPj4gSmltPiBhY3R1 YWxseSB0aGUgaW50ZW50IHdhcyB0aGF0IGJvdGggcGFja2V0LW9yaWVudGVkIGFuZCANCj4+IEpp bT4gYXBwbGljYXRpb24NCj4+IGxheWVyIHNlcnZpY2UgZnVuY3Rpb25zIGFyZSB3aXRoaW4gc2Nv cGUuIEl0IGlzIGNlcnRhaW5seSBwb3NzaWJsZSB0byANCj4+IGVudmlzYWdlIHNlcnZpY2UgY2hh aW5zIHRoYXQgaW5jb3Jwb3JhdGUgYm90aCBvZiB0aGVzZSB0eXBlcyBvZiANCj4+IGZ1bmN0aW9u cyBhbmQgd2Ugc2hvdWxkIG5vdCByZXN0cmljdCBvdXIgd29yayB0byBvbmUgb3IgdGhlIG90aGVy IGJ1dCANCj4+IHJhdGhlciBsZXQgdGhlIHVzZSBjYXNlcyBkaWN0YXRlLg0KPj4NCj4+DQo+Pj4g Q2FuIHdlIGNhcHR1cmUgdGhhdCBpZGVhIGluIGEgZGVmaW5pdGlvbj8NCj4+Pg0KPj4+IEkgbm90 ZSB0aGF0IGl0IGlzIGhpbnRlZCBhdCBpbiBQYXVsIGFuZCBKaW0ncyBCb0YgcHJvcG9zYWwgdGV4 dCBhbmQgDQo+Pj4gaXQgaXMgdGhlcmUgYmVjYXVzZSBJIGFza2VkIHRoZW0gd2h5IHRoaXMgcHJv cG9zYWwgd2FzIGFpbWVkIGF0IHRoZSANCj4+PiBSb3V0aW5nIEFyZWEgbm90IHRoZSBBcHBsaWNh dGlvbnMgQXJlYS4gVG8gd2hpY2ggdGhlIGFuc3dlciB3YXMgDQo+Pj4gImJlY2F1c2Ugd2Ugd2Fu dCB0byBzdGVlciBwYWNrZXRzIG9yIGZsb3dzIHRvIHRoZSBwbGFjZSB3aGVyZSB0aGUgDQo+Pj4g bmV4dCBzZXJ2aWNlIHdpbGwgYmUgYXBwbGllZC4iDQo+Pj4NCj4+PiBBZHJpYW4NCj4+Pg0KPj4+ IF9fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fDQo+Pj4gbnNj IG1haWxpbmcgbGlzdA0KPj4+IG5zY0BpZXRmLm9yZw0KPj4+IGh0dHBzOi8vd3d3LmlldGYub3Jn L21haWxtYW4vbGlzdGluZm8vbnNjDQo+Pg0KPj4gX19fX19fX19fX19fX19fX19fX19fX19fX19f X19fX19fX19fX19fX19fX19fX18NCj4+IG5zYyBtYWlsaW5nIGxpc3QNCj4+IG5zY0BpZXRmLm9y Zw0KPj4gaHR0cHM6Ly93d3cuaWV0Zi5vcmcvbWFpbG1hbi9saXN0aW5mby9uc2MNCj4gX19fX19f X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX18NCj4gbnNjIG1haWxpbmcg bGlzdA0KPiBuc2NAaWV0Zi5vcmcNCj4gaHR0cHM6Ly93d3cuaWV0Zi5vcmcvbWFpbG1hbi9saXN0 aW5mby9uc2MNCg0KDQotLQ0KIkVzdGEgdmV6IG5vIGZhbGxhcmVtb3MsIERvY3RvciBJbmZpZXJu byINCg0KRHIgRGllZ28gUi4gTG9wZXoNClRlbGVmb25pY2EgSStEDQpodHRwOi8vcGVvcGxlLnRp ZC5lcy9kaWVnby5sb3Blei8NCg0KZS1tYWlsOiBkaWVnb0B0aWQuZXMNClRlbDogICAgKzM0IDkx MyAxMjkgMDQxDQpNb2JpbGU6ICszNCA2ODIgMDUxIDA5MQ0KLS0tLS0tLS0tLS0tLS0tLS0tLS0t LS0tLS0tLS0tLS0tLS0tLS0tLS0NCg0KDQpfX19fX19fX19fX19fX19fX19fX19fX19fX19fX19f Xw0KDQpFc3RlIG1lbnNhamUgc2UgZGlyaWdlIGV4Y2x1c2l2YW1lbnRlIGEgc3UgZGVzdGluYXRh cmlvLiBQdWVkZSBjb25zdWx0YXIgbnVlc3RyYSBwb2zDrXRpY2EgZGUgZW52w61vIHkgcmVjZXBj acOzbiBkZSBjb3JyZW8gZWxlY3Ryw7NuaWNvIGVuIGVsIGVubGFjZSBzaXR1YWRvIG3DoXMgYWJh am8uDQpUaGlzIG1lc3NhZ2UgaXMgaW50ZW5kZWQgZXhjbHVzaXZlbHkgZm9yIGl0cyBhZGRyZXNz ZWUuIFdlIG9ubHkgc2VuZCBhbmQgcmVjZWl2ZSBlbWFpbCBvbiB0aGUgYmFzaXMgb2YgdGhlIHRl cm1zIHNldCBvdXQgYXQ6DQpodHRwOi8vd3d3LnRpZC5lcy9FUy9QQUdJTkFTL2Rpc2NsYWltZXIu YXNweA0KX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX18NCm5z YyBtYWlsaW5nIGxpc3QNCm5zY0BpZXRmLm9yZw0KaHR0cHM6Ly93d3cuaWV0Zi5vcmcvbWFpbG1h bi9saXN0aW5mby9uc2MNCl9fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19f X19fX19fDQpuc2MgbWFpbGluZyBsaXN0DQpuc2NAaWV0Zi5vcmcNCmh0dHBzOi8vd3d3LmlldGYu b3JnL21haWxtYW4vbGlzdGluZm8vbnNjDQo= From Ron_Parker@affirmednetworks.com Wed Jun 26 06:50:51 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 332EE21E8126 for ; Wed, 26 Jun 2013 06:50:49 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.599 X-Spam-Level: X-Spam-Status: No, score=-2.599 tagged_above=-999 required=5 tests=[AWL=-0.001, BAYES_00=-2.599, HTML_MESSAGE=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 6iXRjjqpZKTX for ; Wed, 26 Jun 2013 06:50:41 -0700 (PDT) Received: from hub021-ca-7.exch021.serverdata.net (hub021-ca-7.exch021.serverdata.net [64.78.56.72]) by ietfa.amsl.com (Postfix) with ESMTP id 35D4711E80E6 for ; Wed, 26 Jun 2013 06:50:40 -0700 (PDT) Received: from MBX021-W3-CA-2.exch021.domain.local ([10.254.4.78]) by HUB021-CA-7.exch021.domain.local ([10.254.4.109]) with mapi id 14.03.0123.003; Wed, 26 Jun 2013 06:50:40 -0700 From: Ron Parker To: "mohamed.boucadair@orange.com" , "nsc@ietf.org" Thread-Topic: Comments on draft-boucadair-network-function-chaining-00 Thread-Index: Ac5x6/wxcUX+ReJaT8eNlB4kpr2L6QART59wABBNgzA= Date: Wed, 26 Jun 2013 13:50:39 +0000 Message-ID: References: <94C682931C08B048B7A8645303FDC9F36EDB5C7F8F@PUEXCB1B.nanterre.francetelecom.fr> In-Reply-To: <94C682931C08B048B7A8645303FDC9F36EDB5C7F8F@PUEXCB1B.nanterre.francetelecom.fr> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [50.203.66.100] Content-Type: multipart/alternative; boundary="_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66990DMBX021W3CA2exch_" MIME-Version: 1.0 Subject: Re: [nsc] Comments on draft-boucadair-network-function-chaining-00 X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 13:50:51 -0000 --_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66990DMBX021W3CA2exch_ Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Thanks, Med. My followup inline. B.R., Ron From: mohamed.boucadair@orange.com [mailto:mohamed.boucadair@orange.com] Sent: Wednesday, June 26, 2013 2:32 AM To: Ron Parker; nsc@ietf.org Subject: RE: Comments on draft-boucadair-network-function-chaining-00 Dear Ron, Thank you for the review. Please see inline. Cheers, Med De : nsc-bounces@ietf.org [mailto:nsc-bounces@= ietf.org] De la part de Ron Parker Envoy=E9 : mardi 25 juin 2013 23:40 =C0 : nsc@ietf.org Objet : [nsc] Comments on draft-boucadair-network-function-chaining-00 Med, Christian, Thanks for the excellent document and to all that are contributing to this = worthwhile effort. I have included specific comments below and would be p= leased to see them spur some discussion on the thread. Best regards, Ron Parker Affirmed Networks 1.3 Objectives Is it desirable to support unidirectional chains or are only bidirectional = chains considered? [Med] The current text (section 3.3) indicates distinct chains can be appli= ed for inbound and outbound traffic. The direction of the chain is not incl= uded as an attribute of the NLFC Map but this is realized using two chains:= i.e., two NLFC Maps will be installed in the NLFC Policy Table. For bidirectional chains, is NLF chronology always reversed for each dire= ction, or is it arbitrarily controlled? [Med] The current text allows to specify distinct NLFC Map for each directi= on. Are NLF's identified by IP address, FQDN, or both? [Med] I guess your question is about locators not identifiers. The current = text says both can be supported but we recommend the use of an IP address s= ince no resolution capability would be needed in the NLF-enabled node. Reso= lution capabilities (and advanced TE functions) can be enabled only at the = PDP (Policy Decision Point); no need to overload NLF-enabled nodes with suc= h features. [Ron] Yes, I meant locators. Thanks for the clarification. Is it an objective for NLFC to explicitly handle load balancing towards mul= tiply-instantiated NLF's , or is load balancing in this regard considered t= o be part of each NLF? [Med] Yes. The current text says: o Advanced features (e.g., load balancing objectives) are also policy-based. Invoked enforcement points are not aware of such objectives. Is it an objective for NLFC to determine the availability of an NLF? If s= o, are there policies regarding what to do when an NLF becomes unavailable? [Med] Good point. FYI, we are currently working on a OSPF extension for thi= s purpose. We will share the text soon. [Ron] I was thinking more along the lines of ping or BFD to determine if th= e node is reachable. For example, an NLF realizing Analytics may be deeme= d non-critical. If it goes down, the desired behavior may be to prune it = from the chain. Another aspect is load balancing. If there are, for ex= ample, 4 addressable entities provisioned to collectively realize an NLF, i= f one goes down all subsequent traffic should be steered to the remaining 3= . 1.4 Assumptions All NLFs are reachable via standard IP routing techniques (i.e., no assumpt= ions regarding non-IP SDN protocols or techniques)? [Med] How FIBs and RIBs are populated is orthogonal to the overall procedur= e. [Ron] I'm fine with that answer. My thought here was to point out that te= chniques like OpenFlow are not presumed :). 1.5 Rationale Suggest comment about fragmentation be moved to Objectives section. [Med] I will see how to update the text. [Ron]. Thanks. Also, "avoid" fragmentation may be described better as "= minimize" fragmentation. 1.6 Generic Requirements Whether path stretch occurs or not is an administrative matter, assuming re= asonably correct IP routing topologies. [Med] Agree. This is deployment-specific. 3.3 Building NLF Maps Does an NLF map specify the set of NLF's, only, or does it specify the set = of NLF's and the policy to be applied by each. For example, assume that F= irewall-A and NAT64-A are distinct instances of NLFs. Is {Firewall-A-prof= ile-1, NAT64-A-profile-1} distinct from {Firewall-A-profile-2, Nat64-A-prof= ile-1}. Or, alternatively, is the policy conveyed outside of the map? [Med] I'm not sure I get the point. What do you mean by "profile" here? [Ron] By profile, I mean a set of local differentiated policies. For exam= ple, an NLF realizing DPI and traffic enforcement may wish to block peer-to= -peer protocols for some group of users and block voip protocols for some o= ther group of users. Can the PDP facilitate this by conveying to the NLF = whether the traffic belongs to the first group of users or the second? 4.1 NLFC Boundary Node I think the definition of ingress/egress here is problematic as it seems to= be implying topological direction (i.e., uplink and downlink in a mobile e= nvironment). If, however, the NLFC domain is thought of as residing in be= tween 2 non-NLFC domains (or segments of the same domain), then there is bo= th ingress and egress at both ends of the chain. A more descriptive appro= ach may be to use terms such as "West-ingress", "East-egress", "East-ingres= s", "West-egress". In a wireless domain, perhaps "north" and "south" wou= ld be more intuitive, but I think the connotation is more limiting in a gen= eral sense. [Med] Each NLFC Boundary Node behaves in both the ingress and egress roles.= It behaves as "ingress" for traffic entering the domain and as "egress" fo= r traffic existing the domain. [Ron] Ok. This connotation of topological direction may have been my mis= reading of the text. 4.3 NLFC Ingress Node As an ingress node which connects to the non NLFC domain, when would an NLF= C Map Index be present? Also, this implies an encapsulation technique tha= t has not been previously introduced in the document. [Med] Just like setting/re-setting DSCP values at the boundaries of a domai= n, the NLFC Map Index can be set by any "non-legitimate" node. This is even= a basic attack to disturb a service or try granting access to non-authoriz= ed services. This is why we recommend to enforce a basic policy to not trus= t any NLFC Map Index received from adjacent domains. [Ron] My comment was partly editorial - the text describes procedures rel= ated to encapsulation prior to introducing the concept of encapsulation. 5.1.1 NLFC Map Index If the individual entries in an NLFC Map indicate only the identity of the = NLF, then 8 or 16 bits to identify which NLFC Map would be sufficient. If= , however, the entries in an NLFC Map indicate NLF identity+policy, then 16= or 32 bits would be more appropriate. [Med] The policy is not signaled in packets; only the index of the NLFC Map= is. [Ron] I meant policy that is locally significant to the NLF. See my foll= owup to 3.3, above. 5.1.3 Where to store NLFC Map Index in a packet? Consider usage of GRE key. Advantage is that in many cases, encapsulation= is required and GRE is a standard encapsulation technique which provides a= 32-bit general purpose field. Disadvantage is that some systems may requ= ire each unique GRE key to be provisioned as an IP interface. [Med] Will add a new entry for it. Another inconvenient I see is for cases = where encapsulation is not needed at all. This is typically when the next N= LF node is an adjacent node (see Section 6.2). [Ron] Depending on your views on conveyance of NLF-significant policy from= the PDP towards the NLF (see 3.3, above), encapsulation may be desirable e= ven when on the same L3 subnet. --_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66990DMBX021W3CA2exch_ Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

Thanks, Med.

 

My followup inline.

 

B.R.,

Ron<= /p>

 

From: mohamed.boucadair@orange.com [mailto:mo= hamed.boucadair@orange.com]
Sent: Wednesday, June 26, 2013 2:32 AM
To: Ron Parker; nsc@ietf.org
Subject: RE: Comments on draft-boucadair-network-function-chaining-0= 0

 

Dear Ron,

 

Thank you for the review.

Please see inline.

 

Cheers,

Med

 

De : nsc-bounces@ietf.org [mailto:nsc-bounces@ietf.org] De la part de Ron Parker
Envoy=E9 : mardi 25 juin 2013 23:40
=C0 : nsc@ietf.org
Objet : [nsc] Comments on draft-boucadair-network-function-chai= ning-00

 

Med, Christian,

 

Thanks for the excellent document and to all that ar= e contributing to this worthwhile effort.   I have included speci= fic comments below and would be pleased to see them spur some discussion on= the thread.

 

Best regards,

 

Ron Parker

Affirmed Networks

 

 

 

1.3 Objectives

Is it desirable to support unidirectional chains or = are only bidirectional chains considered? 

[Med] The current text (section 3.3) indicat= es distinct chains can be applied for inbound and outbound traffic. The dir= ection of the chain is not included as an attribute of the NLFC Map but this is realized using two chains: i.e., two NLFC Maps= will be installed in the NLFC Policy Table.

 

  For bidirectional chains, is NLF chronology a= lways reversed for each direction, or is it arbitrarily  controlled?

[Med] The current text allows to specify dis= tinct NLFC Map for each direction.

 

Are NLF’s identified by IP address, FQDN, or b= oth?

[Med] I guess your question is about locator= s not identifiers. The current text says both can be supported but we recom= mend the use of an IP address since no resolution capability would be needed in the NLF-enabled node. Resolution capabilitie= s (and advanced TE functions) can be enabled only at the PDP (Policy Decisi= on Point); no need to overload NLF-enabled nodes with such features.

[Ron] Yes, I meant locators.   Thanks for = the clarification.

 

 

Is it an objective for NLFC to explicitly handle loa= d balancing towards multiply-instantiated NLF’s , or is load balancin= g in this regard considered to be part of each NLF?

[Med] Yes. The current text says:=

 

   o  Advanced features (e.g., load balanci= ng objectives) are also

      policy-based.  Invoked= enforcement points are not aware of such

      objectives.

 

 

Is it an objective for NLFC to determine the availab= ility of an NLF?   If so, are there policies regarding what to do= when an NLF becomes unavailable?

[Med] Good point. FYI, we are currently work= ing on a OSPF extension for this purpose. We will share the text soon.=

[Ron] I was thinking more along the lines of ping or= BFD to determine if the node is reachable.   For example, an NLF= realizing Analytics may be deemed non-critical.   If it goes dow= n, the desired behavior may be to prune it from the chain.   = ; Another aspect is load balancing.   If there are, for example, 4= addressable entities provisioned to collectively realize an NLF, if one go= es down all subsequent traffic should be steered to the remaining 3.

 

1.4 Assumptions

All NLFs are reachable via standard IP routing techn= iques (i.e., no assumptions regarding non-IP SDN protocols or techniques)?<= o:p>

[Med] How FIBs and RIBs are populated is ort= hogonal to the overall procedure.

[Ron] I’m fine with that answer.   M= y thought here was to point out that techniques like OpenFlow are not presu= med J.

 

1.5 Rationale

Suggest comment about fragmentation be moved to Obje= ctives section.

[Med] I will see how to update the text.

[Ron].   Thanks.   Also, “= avoid” fragmentation may be described better as “minimize”= ; fragmentation.

 

1.6 Generic Requirements

Whether path stretch occurs or not is an administrat= ive matter, assuming reasonably correct IP routing topologies.

[Med] Agree. This is deployment-specific.

 

3.3 Building NLF Maps

Does an NLF map specify the set of NLF’s, only= , or does it specify the set of NLF’s and the policy to be applied by= each.   For example, assume that Firewall-A and NAT64-A are dist= inct instances of NLFs.   Is {Firewall-A-profile-1, NAT64-A-profi= le-1} distinct from {Firewall-A-profile-2, Nat64-A-profile-1}.   Or, a= lternatively, is the policy conveyed outside of the map?

[Med] I’m not sure I get the point. Wh= at do you mean by “profile” here?

[Ron] By profile, I mean a set of local differentiat= ed policies.   For example, an NLF realizing DPI and traffic enfo= rcement may wish to block peer-to-peer protocols for some group of users an= d block voip protocols for some other group of users.   Can the PDP facilitate this by conveying to the NLF = whether the traffic belongs to the first group of users or the second?=

 

4.1  NLFC Boundary Node

I think the definition of ingress/egress here is pro= blematic as it seems to be implying topological direction (i.e., uplink and= downlink in a mobile environment).   If, however, the NLFC domai= n is thought of as residing in between 2 non-NLFC domains (or segments of the same domain), then there is both ingress and e= gress at both ends of the chain.   A more descriptive approach ma= y be to use terms such as “West-ingress”, “East-egress= 221;, “East-ingress”, “West-egress”.  &nb= sp; In a wireless domain, perhaps “north” and “south” would be more intuitive, but I= think the connotation is more limiting in a general sense.

[Med] Each NLFC Boundary Node behaves in bot= h the ingress and egress roles. It behaves as “ingress” for tra= ffic entering the domain and as “egress” for traffic existing the domain.

[Ron]  Ok.   This connotation of topo= logical direction may have been my misreading of the text.

 

4.3 NLFC Ingress Node

As an ingress node which connects to the non NLFC do= main, when would an NLFC Map Index be present?   Also, this impli= es an encapsulation technique that has not been previously introduced in th= e document.

[Med] Just like setting/re-setting DSCP valu= es at the boundaries of a domain, the NLFC Map Index can be set by any R= 20;non-legitimate” node. This is even a basic attack to disturb a service or try granting access to non-authorized services. Th= is is why we recommend to enforce a basic policy to not trust any NLFC Map = Index received from adjacent domains.

[Ron]   My comment was partly editorial &#= 8211; the text describes procedures related to encapsulation prior to intro= ducing the concept of encapsulation.

 

5.1.1 NLFC Map Index

If the individual entries in an NLFC Map indicate on= ly the identity of the NLF, then 8 or 16 bits to identify which NLFC Map wo= uld be sufficient.   If, however, the entries in an NLFC Map indi= cate NLF identity+policy, then 16 or 32 bits would be more appropriate.

[Med] The policy is not signaled in packets;= only the index of the NLFC Map is.  

[Ron]  I meant policy that is locally significa= nt to the NLF.   See my followup to 3.3, above.

 

5.1.3 Where to store NLFC Map Index in a packet?

Consider usage of GRE key.   Advantage is = that in many cases, encapsulation is required and GRE is a standard encapsu= lation technique which provides a 32-bit general purpose field.  = Disadvantage is that some systems may require each unique GRE key to be provisioned as an IP interface.

[Med] Will add a new entry for it. Another i= nconvenient I see is for cases where encapsulation is not needed at all. Th= is is typically when the next NLF node is an adjacent node (see Section 6.2).

[Ron]  Depending on your views on conveyance of= NLF-significant policy from the PDP towards the NLF (see 3.3, above), enca= psulation may be desirable even when on the same L3 subnet.

 

 

--_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66990DMBX021W3CA2exch_-- From diego@tid.es Wed Jun 26 06:55:32 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id B06E011E80E6 for ; Wed, 26 Jun 2013 06:55:21 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id DMkci+LIyJHN for ; Wed, 26 Jun 2013 06:55:15 -0700 (PDT) Received: from correo-bck.tid.es (correo-bck.tid.es [195.235.93.200]) by ietfa.amsl.com (Postfix) with ESMTP id B43D421E8140 for ; Wed, 26 Jun 2013 06:55:14 -0700 (PDT) Received: from sbrightmailg02.hi.inet (Sbrightmailg02.hi.inet [10.95.78.105]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTP id <0MP000J9H6O0UW@tid.hi.inet> for nsc@ietf.org; Wed, 26 Jun 2013 15:55:12 +0200 (MEST) Received: from vanvan (vanvan.hi.inet [10.95.78.49]) by sbrightmailg02.hi.inet (Symantec Messaging Gateway) with SMTP id 8B.48.02911.0C2FAC15; Wed, 26 Jun 2013 15:55:12 +0200 (CEST) Received: from correo.tid.es (mailhost.hi.inet [10.95.64.100]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTPS id <0MP000J996NZUW@tid.hi.inet> for nsc@ietf.org; Wed, 26 Jun 2013 15:55:12 +0200 (MEST) Received: from EX10-MB2-MAD.hi.inet ([169.254.2.38]) by EX10-HTCAS8-MAD.hi.inet ([fe80::41c8:e965:8a6:de67%11]) with mapi id 14.02.0328.009; Wed, 26 Jun 2013 15:55:11 +0200 Date: Wed, 26 Jun 2013 13:54:08 +0000 From: "Diego R. Lopez" In-reply-to: X-Originating-IP: [10.95.64.115] To: Ron Parker Message-id: Content-id: MIME-version: 1.0 Content-type: text/plain; charset=utf-8 Content-language: en-US Content-transfer-encoding: base64 Accept-Language: en-US, es-ES Thread-topic: [nsc] What is a service? Why is NSC not SC? Thread-index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewAAIKxYAABiYCAAAeHsEAAISBAQAAbK0RAAAVvSOAAAFzfAA= X-AuditID: 0a5f4e69-b7f118e000000b5f-4e-51caf2c068cd X-MS-Has-Attach: X-MS-TNEF-Correlator: X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFmphkeLIzCtJLcpLzFFi42Lhivcz1D3w6VSgQcMpcYvj+/azOTB6LFny kymAMYrLJiU1J7MstUjfLoErY+XJn8wFr4Qq5m15zNjAuEaoi5GTQ0LAROL3j6fsELaYxIV7 69lAbCGB7YwSF09WQtjfGSWu7VDrYuQCsjcwSvzeNwusgUVAVeLQ+pcsIDYbkP2o+TdYXFjA QmLbzX9MIDanQKzElbmf2SAWKEj8OfcYrF5EwEDi9edTrCBDmQUmMUk8fXiJFSTBK+At8fXe O7BBzAJmEtfv3GSEiAtK/Jh8D6iZAyiuLjFlSi5EibhEc+tNFghbUWLaogawckYBWYl38+ez gpSLCFhKNHySg1ibInFpyiNmiHMEJJbsOQ9li0q8fPyPFeLHdcwSLzbOZZrAKDELyRWzkFwx C+GKWUiumIXkigWMrKsYxYqTijLTM0pyEzNz0g2M9DIy9TLzUks2MUJiLnMH4/KdKocYBTgY lXh4FbaeDBRiTSwrrsw9xCjBwawkwvtm/qlAId6UxMqq1KL8+KLSnNTiQ4xMHJxSDYx2Dqf4 OFap27mGHDaNdZLYmFuUExh2obFYZuLbCsl3re+7Va9ey1Y2jfIR7Stp/3yHq23dUbbJnmEf bOzN73qUfQ5Y+W+F9eV+oeDSdXpTubbdmLliTvKlZTX7HEOj2uVcwg9Eu08St7hY655UZviC JW3T1yXnDqZ/Cv+66hz38vVsipN+HVJiKc5INNRiLipOBAAe6/LClwIAAA== References: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> <6EB34CB5D82C4645B826C56144826EA97DF0AD29@szxeml560-mbx.china.huawei.com> Cc: "nsc@ietf.org" , "Jim Guichard \(jguichar\)" , "Hongyu Li \(Julio\)" , " " , "adrian@olddog.co.uk" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 13:55:32 -0000 SGkgUm9uLA0KDQpPbiAyNiBKdW4gMjAxMywgYXQgMTU6MTMgLCBSb24gUGFya2VyIHdyb3RlOg0K Pg0KPiBbc25pcF0gQW5vdGhlciBwb2ludCBJIHdhbm5hIG1ha2UgaXMgdGhhdCBJTUhPLCBvbiBr ZXkgY2hhcmFjdGVyIG9mIE5TQyBpcyB0cmFmZmljIGlzIGZvcndhcmRlZCBwZXIgc2VydmljZSBj aGFyYWN0ZXJpc3RpY3Mgb3RoZXIgdGhhbiBkZXN0aW5hdGlvbiBhZGRyZXNzZXMuIFRoaXMgaXMg dGhlIGtleSB2YWx1ZSBOU0MgV0csIGlmIHdpbGwgYmUgZm9ybWVkLCBzaG91bGQgd29yayBvbi4g U2VydmljZSBjaGFyYWN0ZXJpc3RpY3MgbWF5IGluY2x1ZGUgdGhlIG5ldHdvcmsgc2VydmljZSBh IHVzZSBoYXMgc3Vic2NyaWJlZCB0bywgZS5nLiBwYXJlbnRhbCBjb250cm9sLCBmaXJld2FsbC4g VGhhdCBtZWFucyBhIHBhY2tldCBpcyBmb3J3YXJkZWQgaW50byBhIHNlcnZpY2UgY2hhaW4sIHdo ZW4gYSBjbGFzc2lmaWVyIG1hdGNoZXMgc29tZSBydWxlcyBvZiBzZXJ2aWNlIGNoYXJhY3Rlcmlz dGljcywgd2hpY2ggbWF5IGJlIHJlZmxlY3RlZCBieSBhY2Nlc3MgcG9ydHMsIG9yIHNlZ21lbnRz IG9yIHR1cGxlcyBpbiBhIHBhY2tldC4NCj4NCj4gW1Jvbl06ICBJIHZpZXcgTlNDIGFzIGEgc3Bl Y2lmaWMgYXBwbGljYXRpb24gb2YgcG9saWN5IGJhc2VkIGZvcndhcmRpbmcsIHdoZXJlIHRoZSBw b2xpY3kgaXMgYmFzZWQgb24gYW55IG9yIGFsbCBvZiB0aGUgZmllbGRzIGluIHRoZSBJUCBoZWFk ZXIuICAgVGhlIGVuZC10by1lbmQgbW9kZWwgb2YgY2xhc3NpZmllci1iYXNlZCBub2RlcyBhbmQg bm9uLWNsYXNzaWZpZXItYmFzZWQgbm9kZXMgbWVhbnMgdGhhdCB0aGVyZSBhcmUgMiBkaXN0aW5j dCBzZXRzIG9mIHByb2NlZHVyZXMgdG8gZGV0ZXJtaW5lIGFuZCBhcHBseSB0aGUgcG9saWN5Lg0K DQoNCkknZCBwcmVmZXIgdG8gdGhpbmsgdGhhdCBpdCBpcyBub3QgbmVjZXNzYXJpbHkgbGltaXRl ZCB0byB0aGUgZmllbGRzIGluIHRoZSBJUCBoZWFkZXIuIFRvIHN0YXJ0IHdpdGgsIHdlIGNvdWxk IHRoaW5rIGluIFRDUC9VRFAgaGVhZGVycyBvciBMMiBvbmVzLiBPciBldmVuIHRoaW5rIGFib3V0 IGFwcGxpY2F0aW9uIHN0dWZmLCBsaWtlIEhUVFAgaGVhZGVycy4uLg0KDQpCZSBnb29kZSwNCg0K LS0NCiJFc3RhIHZleiBubyBmYWxsYXJlbW9zLCBEb2N0b3IgSW5maWVybm8iDQoNCkRyIERpZWdv IFIuIExvcGV6DQpUZWxlZm9uaWNhIEkrRA0KaHR0cDovL3Blb3BsZS50aWQuZXMvZGllZ28ubG9w ZXovDQoNCmUtbWFpbDogZGllZ29AdGlkLmVzDQpUZWw6ICAgICszNCA5MTMgMTI5IDA0MQ0KTW9i aWxlOiArMzQgNjgyIDA1MSAwOTENCi0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0t LS0tLS0tDQoNCg0KX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX18NCg0KRXN0ZSBtZW5z YWplIHNlIGRpcmlnZSBleGNsdXNpdmFtZW50ZSBhIHN1IGRlc3RpbmF0YXJpby4gUHVlZGUgY29u c3VsdGFyIG51ZXN0cmEgcG9sw610aWNhIGRlIGVudsOtbyB5IHJlY2VwY2nDs24gZGUgY29ycmVv IGVsZWN0csOzbmljbyBlbiBlbCBlbmxhY2Ugc2l0dWFkbyBtw6FzIGFiYWpvLg0KVGhpcyBtZXNz YWdlIGlzIGludGVuZGVkIGV4Y2x1c2l2ZWx5IGZvciBpdHMgYWRkcmVzc2VlLiBXZSBvbmx5IHNl bmQgYW5kIHJlY2VpdmUgZW1haWwgb24gdGhlIGJhc2lzIG9mIHRoZSB0ZXJtcyBzZXQgb3V0IGF0 Og0KaHR0cDovL3d3dy50aWQuZXMvRVMvUEFHSU5BUy9kaXNjbGFpbWVyLmFzcHgNCg== From dmm@1-4-5.net Wed Jun 26 06:56:57 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 62E5021F9EB6 for ; Wed, 26 Jun 2013 06:56:57 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: 0.479 X-Spam-Level: X-Spam-Status: No, score=0.479 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, FH_RELAY_NODNS=1.451, FM_FORGED_GMAIL=0.622, RCVD_IN_PBL=0.905, RDNS_NONE=0.1] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id e9SKyaRIt-8B for ; Wed, 26 Jun 2013 06:56:53 -0700 (PDT) Received: from mail-oa0-x22d.google.com (mail-oa0-x22d.google.com [IPv6:2607:f8b0:4003:c02::22d]) by ietfa.amsl.com (Postfix) with ESMTP id E95A221F9E7E for ; Wed, 26 Jun 2013 06:56:52 -0700 (PDT) Received: by mail-oa0-f45.google.com with SMTP id j1so14701720oag.4 for ; Wed, 26 Jun 2013 06:56:38 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:x-originating-ip:in-reply-to:references:date :message-id:subject:from:to:cc:content-type :content-transfer-encoding:x-gm-message-state; bh=bSJH+O3A0Sv1RR/C57cWme4pKPBQdU5OImKxovN+MhM=; b=MuToIL9/EEBINs3mUBzcojrsRozMkdG4Gc1+o+EHSd3XrwjfhigDIUPwfdXCcC4aAd i1FAtudmGdhxWv5UrQJHMG23W+wkrb+d398qcjxXeUdhzd0ok2nKpmmcEMCzAA00BDEw MklT1pHwLFgST4LEGhVorO2Upw95XYxijRhIZPyC3KOkBatvaIDsJ5D+OljdH1HAlczo WJjomG58GDHqFKFoWeG2VDbmkL2t9QlWgHxf9SJMngKE0UhEhKXauQVtTRL2Wav7a8NA SsICc8GmEUn7tNOCgqwpq+HMlO1rKImoRAPML6Yq0qnZbBZlNSTdPHSvx2jgWUVsauN5 AW4Q== MIME-Version: 1.0 X-Received: by 10.182.171.7 with SMTP id aq7mr537416obc.103.1372254998180; Wed, 26 Jun 2013 06:56:38 -0700 (PDT) Received: by 10.182.114.34 with HTTP; Wed, 26 Jun 2013 06:56:38 -0700 (PDT) X-Originating-IP: [64.134.225.12] In-Reply-To: References: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> <6EB34CB5D82C4645B826C56144826EA97DF0AD29@szxeml560-mbx.china.huawei.com> Date: Wed, 26 Jun 2013 06:56:38 -0700 Message-ID: From: David Meyer To: "Diego R. Lopez" Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable X-Gm-Message-State: ALoCoQmAIu/4ljvLpuoOxey9I2PaXMfv7JGglcQ8MX3HOShl+kwJjnU1AO6A65LMmv6Hi8FOlPMe Cc: "nsc@ietf.org" , "Jim Guichard \(jguichar\)" , "Hongyu Li \(Julio\)" , Ron Parker , "adrian@olddog.co.uk" , " " Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 13:56:57 -0000 On Wed, Jun 26, 2013 at 6:54 AM, Diego R. Lopez wrote: > Hi Ron, > > On 26 Jun 2013, at 15:13 , Ron Parker wrote: >> >> [snip] Another point I wanna make is that IMHO, on key character of NSC = is traffic is forwarded per service characteristics other than destination = addresses. This is the key value NSC WG, if will be formed, should work on.= Service characteristics may include the network service a use has subscrib= ed to, e.g. parental control, firewall. That means a packet is forwarded in= to a service chain, when a classifier matches some rules of service charact= eristics, which may be reflected by access ports, or segments or tuples in = a packet. >> >> [Ron]: I view NSC as a specific application of policy based forwarding,= where the policy is based on any or all of the fields in the IP header. = The end-to-end model of classifier-based nodes and non-classifier-based nod= es means that there are 2 distinct sets of procedures to determine and appl= y the policy. > > > I'd prefer to think that it is not necessarily limited to the fields in t= he IP header. To start with, we could think in TCP/UDP headers or L2 ones. = Or even think about application stuff, like HTTP headers... Certainly agree with this...--dmm > > Be goode, > > -- > "Esta vez no fallaremos, Doctor Infierno" > > Dr Diego R. Lopez > Telefonica I+D > http://people.tid.es/diego.lopez/ > > e-mail: diego@tid.es > Tel: +34 913 129 041 > Mobile: +34 682 051 091 > ----------------------------------------- > > > ________________________________ > > Este mensaje se dirige exclusivamente a su destinatario. Puede consultar = nuestra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el e= nlace situado m=E1s abajo. > This message is intended exclusively for its addressee. We only send and = receive email on the basis of the terms set out at: > http://www.tid.es/ES/PAGINAS/disclaimer.aspx > _______________________________________________ > nsc mailing list > nsc@ietf.org > https://www.ietf.org/mailman/listinfo/nsc From Ron_Parker@affirmednetworks.com Wed Jun 26 06:59:25 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 8FA5021F9F75 for ; Wed, 26 Jun 2013 06:59:25 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.599 X-Spam-Level: X-Spam-Status: No, score=-2.599 tagged_above=-999 required=5 tests=[AWL=0.000, BAYES_00=-2.599] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ljlDUTJ7UABD for ; Wed, 26 Jun 2013 06:59:20 -0700 (PDT) Received: from hub021-ca-1.exch021.serverdata.net (hub021-ca-1.exch021.serverdata.net [64.78.22.168]) by ietfa.amsl.com (Postfix) with ESMTP id 4327721F9EE2 for ; Wed, 26 Jun 2013 06:59:20 -0700 (PDT) Received: from MBX021-W3-CA-2.exch021.domain.local ([10.254.4.78]) by HUB021-CA-1.exch021.domain.local ([10.254.4.30]) with mapi id 14.03.0123.003; Wed, 26 Jun 2013 06:59:02 -0700 From: Ron Parker To: "Diego R. Lopez" Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrew//+rwICAADExAIAA8PYAgAQjwYD//Bu+0P/3g9MggBF6zgCAAHTVkA== Date: Wed, 26 Jun 2013 13:59:02 +0000 Message-ID: References: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> <6EB34CB5D82C4645B826C56144826EA97DF0AD29@szxeml560-mbx.china.huawei.com> In-Reply-To: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [50.203.66.100] Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: base64 MIME-Version: 1.0 Cc: "nsc@ietf.org" , "Jim Guichard \(jguichar\)" , "Hongyu Li \(Julio\)" , " " , "adrian@olddog.co.uk" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 13:59:25 -0000 SGksIERpZWdvLg0KDQpUaGFua3MgZm9yIHRoZSByZXNwb25zZS4gICBDb21tZW50IGJlbG93Lg0K DQpCUiwNClJvbg0KDQoNCltzbmlwXSBJJ2QgcHJlZmVyIHRvIHRoaW5rIHRoYXQgaXQgaXMgbm90 IG5lY2Vzc2FyaWx5IGxpbWl0ZWQgdG8gdGhlIGZpZWxkcyBpbiB0aGUgSVAgaGVhZGVyLiBUbyBz dGFydCB3aXRoLCB3ZSBjb3VsZCB0aGluayBpbiBUQ1AvVURQIGhlYWRlcnMgb3IgTDIgb25lcy4g T3IgZXZlbiB0aGluayBhYm91dCBhcHBsaWNhdGlvbiBzdHVmZiwgbGlrZSBIVFRQIGhlYWRlcnMu Li4NCg0KW1Jvbl0gQWdyZWUgY29tcGxldGVseS4gICBOU0MgaXMgYSBzcGVjaWZpYyBhcHBsaWNh dGlvbiBvZiBwb2xpY3kgYmFzZWQgZm9yd2FyZGluZyB3aGVyZSB0aGUgcG9saWN5IGlzIGJhc2Vk IG9uIGNsYXNzaWZpY2F0aW9uIG9mIHRoZSBwYWNrZXRzLCBpbmNsdWRpbmcgbGF5ZXJzIDIgdGhy b3VnaCA3LiAgICAgQnkgc3RhdGluZyB0aGF0IHRoZSBOTEYncyBhcmUgbG9jYXRlZCB2aWEgdGhl aXIgSVAgYWRkcmVzc2VzIGFuZC9vciBGUUROJ3MsIHdlIGRvbid0IHJlcXVpcmUgb3IgcHJlY2x1 ZGUgYW55IHBhcnRpY3VsYXIgZm9yd2FyZGluZyBwYXJhZGlnbSBiZXR3ZWVuIHRoZSBjbGFzc2lm aWVyIG5vZGUgYW5kIHRoZSBub24tY2xhc3NpZmllciBub2RlLg0KDQoNCi0tLS0tT3JpZ2luYWwg TWVzc2FnZS0tLS0tDQpGcm9tOiBEaWVnbyBSLiBMb3BleiBbbWFpbHRvOmRpZWdvQHRpZC5lc10g DQpTZW50OiBXZWRuZXNkYXksIEp1bmUgMjYsIDIwMTMgOTo1NCBBTQ0KVG86IFJvbiBQYXJrZXIN CkNjOiBIb25neXUgTGkgKEp1bGlvKTsgPG1vaGFtZWQuYm91Y2FkYWlyQG9yYW5nZS5jb20+IDxt b2hhbWVkLmJvdWNhZGFpckBvcmFuZ2UuY29tPjsgbnNjQGlldGYub3JnOyBKaW0gR3VpY2hhcmQg KGpndWljaGFyKTsgYWRyaWFuQG9sZGRvZy5jby51aw0KU3ViamVjdDogUmU6IFtuc2NdIFdoYXQg aXMgYSBzZXJ2aWNlPyBXaHkgaXMgTlNDIG5vdCBTQz8NCg0KSGkgUm9uLA0KDQpPbiAyNiBKdW4g MjAxMywgYXQgMTU6MTMgLCBSb24gUGFya2VyIHdyb3RlOg0KPg0KPiBbc25pcF0gQW5vdGhlciBw b2ludCBJIHdhbm5hIG1ha2UgaXMgdGhhdCBJTUhPLCBvbiBrZXkgY2hhcmFjdGVyIG9mIE5TQyBp cyB0cmFmZmljIGlzIGZvcndhcmRlZCBwZXIgc2VydmljZSBjaGFyYWN0ZXJpc3RpY3Mgb3RoZXIg dGhhbiBkZXN0aW5hdGlvbiBhZGRyZXNzZXMuIFRoaXMgaXMgdGhlIGtleSB2YWx1ZSBOU0MgV0cs IGlmIHdpbGwgYmUgZm9ybWVkLCBzaG91bGQgd29yayBvbi4gU2VydmljZSBjaGFyYWN0ZXJpc3Rp Y3MgbWF5IGluY2x1ZGUgdGhlIG5ldHdvcmsgc2VydmljZSBhIHVzZSBoYXMgc3Vic2NyaWJlZCB0 bywgZS5nLiBwYXJlbnRhbCBjb250cm9sLCBmaXJld2FsbC4gVGhhdCBtZWFucyBhIHBhY2tldCBp cyBmb3J3YXJkZWQgaW50byBhIHNlcnZpY2UgY2hhaW4sIHdoZW4gYSBjbGFzc2lmaWVyIG1hdGNo ZXMgc29tZSBydWxlcyBvZiBzZXJ2aWNlIGNoYXJhY3RlcmlzdGljcywgd2hpY2ggbWF5IGJlIHJl ZmxlY3RlZCBieSBhY2Nlc3MgcG9ydHMsIG9yIHNlZ21lbnRzIG9yIHR1cGxlcyBpbiBhIHBhY2tl dC4NCj4NCj4gW1Jvbl06ICBJIHZpZXcgTlNDIGFzIGEgc3BlY2lmaWMgYXBwbGljYXRpb24gb2Yg cG9saWN5IGJhc2VkIGZvcndhcmRpbmcsIHdoZXJlIHRoZSBwb2xpY3kgaXMgYmFzZWQgb24gYW55 IG9yIGFsbCBvZiB0aGUgZmllbGRzIGluIHRoZSBJUCBoZWFkZXIuICAgVGhlIGVuZC10by1lbmQg bW9kZWwgb2YgY2xhc3NpZmllci1iYXNlZCBub2RlcyBhbmQgbm9uLWNsYXNzaWZpZXItYmFzZWQg bm9kZXMgbWVhbnMgdGhhdCB0aGVyZSBhcmUgMiBkaXN0aW5jdCBzZXRzIG9mIHByb2NlZHVyZXMg dG8gZGV0ZXJtaW5lIGFuZCBhcHBseSB0aGUgcG9saWN5Lg0KDQoNCkknZCBwcmVmZXIgdG8gdGhp bmsgdGhhdCBpdCBpcyBub3QgbmVjZXNzYXJpbHkgbGltaXRlZCB0byB0aGUgZmllbGRzIGluIHRo ZSBJUCBoZWFkZXIuIFRvIHN0YXJ0IHdpdGgsIHdlIGNvdWxkIHRoaW5rIGluIFRDUC9VRFAgaGVh ZGVycyBvciBMMiBvbmVzLiBPciBldmVuIHRoaW5rIGFib3V0IGFwcGxpY2F0aW9uIHN0dWZmLCBs aWtlIEhUVFAgaGVhZGVycy4uLg0KDQpCZSBnb29kZSwNCg0KLS0NCiJFc3RhIHZleiBubyBmYWxs YXJlbW9zLCBEb2N0b3IgSW5maWVybm8iDQoNCkRyIERpZWdvIFIuIExvcGV6DQpUZWxlZm9uaWNh IEkrRA0KaHR0cDovL3Blb3BsZS50aWQuZXMvZGllZ28ubG9wZXovDQoNCmUtbWFpbDogZGllZ29A dGlkLmVzDQpUZWw6ICAgICszNCA5MTMgMTI5IDA0MQ0KTW9iaWxlOiArMzQgNjgyIDA1MSAwOTEN Ci0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tDQoNCg0KX19fX19fX19f X19fX19fX19fX19fX19fX19fX19fX18NCg0KRXN0ZSBtZW5zYWplIHNlIGRpcmlnZSBleGNsdXNp dmFtZW50ZSBhIHN1IGRlc3RpbmF0YXJpby4gUHVlZGUgY29uc3VsdGFyIG51ZXN0cmEgcG9sw610 aWNhIGRlIGVudsOtbyB5IHJlY2VwY2nDs24gZGUgY29ycmVvIGVsZWN0csOzbmljbyBlbiBlbCBl bmxhY2Ugc2l0dWFkbyBtw6FzIGFiYWpvLg0KVGhpcyBtZXNzYWdlIGlzIGludGVuZGVkIGV4Y2x1 c2l2ZWx5IGZvciBpdHMgYWRkcmVzc2VlLiBXZSBvbmx5IHNlbmQgYW5kIHJlY2VpdmUgZW1haWwg b24gdGhlIGJhc2lzIG9mIHRoZSB0ZXJtcyBzZXQgb3V0IGF0Og0KaHR0cDovL3d3dy50aWQuZXMv RVMvUEFHSU5BUy9kaXNjbGFpbWVyLmFzcHgNCg== From jguichar@cisco.com Wed Jun 26 07:17:03 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id BB01F1F0D33 for ; Wed, 26 Jun 2013 07:17:03 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.599 X-Spam-Level: X-Spam-Status: No, score=-10.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id D+Na+rX4hhoJ for ; Wed, 26 Jun 2013 07:16:58 -0700 (PDT) Received: from rcdn-iport-6.cisco.com (rcdn-iport-6.cisco.com [173.37.86.77]) by ietfa.amsl.com (Postfix) with ESMTP id 8041421F9FE2 for ; Wed, 26 Jun 2013 07:16:55 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=2372; q=dns/txt; s=iport; t=1372256215; x=1373465815; h=from:to:cc:subject:date:message-id:references: in-reply-to:content-transfer-encoding:mime-version; bh=KIKf6YuWj8UufhTs/3jv3jfc438P4eoAmo2m3uZtXaE=; b=JX/2cGoie8ywf39e/Da5Q7xWBIhluxCz0PhU5FQbcvAyfF9YmVmvc6gT diifdfDgDJB5IHTbA7EId+fJTIOWN/ErHkpaGfczpvCIg81ZGW4ypQf1p kbIYGCSN2pT/uZoZ5UgwELY/D3PK3uKtKNhDTsN+1qIageG6yUYdWjgW1 0=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AhAFAD33ylGtJV2b/2dsb2JhbABagwkxvjqBCoEEFnSCIwEBAQMBeQULAgEIGC4yJQIEDgWICAYMulyOCQoGfzMHgwJhA4hqjluBKYkGhxaDEYFoCRc X-IronPort-AV: E=Sophos;i="4.87,944,1363132800"; d="scan'208";a="227678222" Received: from rcdn-core-4.cisco.com ([173.37.93.155]) by rcdn-iport-6.cisco.com with ESMTP; 26 Jun 2013 14:16:55 +0000 Received: from xhc-aln-x07.cisco.com (xhc-aln-x07.cisco.com [173.36.12.81]) by rcdn-core-4.cisco.com (8.14.5/8.14.5) with ESMTP id r5QEGsgG026363 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Wed, 26 Jun 2013 14:16:54 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.80]) by xhc-aln-x07.cisco.com ([173.36.12.81]) with mapi id 14.02.0318.004; Wed, 26 Jun 2013 09:16:54 -0500 From: "Jim Guichard (jguichar)" To: "Diego R. Lopez" Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewABC13ID//+4ggP//KpNAgAYtNICAA2WvAIAAremAgAALUgD//7KLOA== Date: Wed, 26 Jun 2013 14:16:53 +0000 Message-ID: References: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> <6EB34CB5D82C4645B826C56144826EA97DF0AD29@szxeml560-mbx.china.huawei.com> , In-Reply-To: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Cc: "nsc@ietf.org" , "adrian@olddog.co.uk" , "Hongyu Li \(Julio\)" , " " , Ron Parker Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 14:17:03 -0000 I think we need to be clear on definitions. I believe your comments to be t= rue (although I would not restrict it to the IP header only) for classifica= tion of traffic that needs to be forwarded into a service chain. Actual for= warding through the service chain is achieved using the headers pushed on t= o the packets based upon the initial or subsequent classification. Sent from my iPhone On Jun 26, 2013, at 9:55 AM, "Diego R. Lopez" wrote: > Hi Ron, >=20 > On 26 Jun 2013, at 15:13 , Ron Parker wrote: >>=20 >> [snip] Another point I wanna make is that IMHO, on key character of NSC = is traffic is forwarded per service characteristics other than destination = addresses. This is the key value NSC WG, if will be formed, should work on.= Service characteristics may include the network service a use has subscrib= ed to, e.g. parental control, firewall. That means a packet is forwarded in= to a service chain, when a classifier matches some rules of service charact= eristics, which may be reflected by access ports, or segments or tuples in = a packet. >>=20 >> [Ron]: I view NSC as a specific application of policy based forwarding,= where the policy is based on any or all of the fields in the IP header. = The end-to-end model of classifier-based nodes and non-classifier-based nod= es means that there are 2 distinct sets of procedures to determine and appl= y the policy. >=20 >=20 > I'd prefer to think that it is not necessarily limited to the fields in t= he IP header. To start with, we could think in TCP/UDP headers or L2 ones. = Or even think about application stuff, like HTTP headers... >=20 > Be goode, >=20 > -- > "Esta vez no fallaremos, Doctor Infierno" >=20 > Dr Diego R. Lopez > Telefonica I+D > http://people.tid.es/diego.lopez/ >=20 > e-mail: diego@tid.es > Tel: +34 913 129 041 > Mobile: +34 682 051 091 > ----------------------------------------- >=20 >=20 > ________________________________ >=20 > Este mensaje se dirige exclusivamente a su destinatario. Puede consultar = nuestra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el e= nlace situado m=E1s abajo. > This message is intended exclusively for its addressee. We only send and = receive email on the basis of the terms set out at: > http://www.tid.es/ES/PAGINAS/disclaimer.aspx From diego@tid.es Wed Jun 26 07:38:25 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6BC3521E80E3 for ; Wed, 26 Jun 2013 07:38:25 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id aCyx2nPKEJue for ; Wed, 26 Jun 2013 07:38:21 -0700 (PDT) Received: from correo-bck.tid.es (correo-bck.tid.es [195.235.93.200]) by ietfa.amsl.com (Postfix) with ESMTP id 3CFFC21E809A for ; Wed, 26 Jun 2013 07:38:20 -0700 (PDT) Received: from sbrightmailg02.hi.inet (Sbrightmailg02.hi.inet [10.95.78.105]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTP id <0MP000J0M8NTUW@tid.hi.inet> for nsc@ietf.org; Wed, 26 Jun 2013 16:38:17 +0200 (MEST) Received: from vanvan (vanvan.hi.inet [10.95.78.49]) by sbrightmailg02.hi.inet (Symantec Messaging Gateway) with SMTP id 19.D9.02911.9DCFAC15; Wed, 26 Jun 2013 16:38:17 +0200 (CEST) Received: from correo.tid.es (mailhost.hi.inet [10.95.64.100]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTPS id <0MP000J0E8NRUW@tid.hi.inet> for nsc@ietf.org; Wed, 26 Jun 2013 16:38:16 +0200 (MEST) Received: from EX10-MB2-MAD.hi.inet ([169.254.2.38]) by EX10-HTCAS8-MAD.hi.inet ([fe80::41c8:e965:8a6:de67%11]) with mapi id 14.02.0328.009; Wed, 26 Jun 2013 16:38:15 +0200 Date: Wed, 26 Jun 2013 14:37:12 +0000 From: "Diego R. Lopez" In-reply-to: X-Originating-IP: [10.95.64.115] To: "Jim Guichard (jguichar)" Message-id: Content-id: <96249479B66E1241A5764EB44A56C143@hi.inet> MIME-version: 1.0 Content-type: text/plain; charset=iso-8859-1 Content-language: en-US Content-transfer-encoding: quoted-printable Accept-Language: en-US, es-ES Thread-topic: [nsc] What is a service? Why is NSC not SC? Thread-index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewAAIKxYAABiYCAAAeHsEAAISBAQAAbK0RAAAVvSOAAAFzfAAAAMIqgAAAvuIA X-AuditID: 0a5f4e69-b7f118e000000b5f-c5-51cafcd972ba X-MS-Has-Attach: X-MS-TNEF-Correlator: X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFmpjkeLIzCtJLcpLzFFi42Lhivcz1L3551SgQfc9Lovj+/azOTB6LFny kymAMYrLJiU1J7MstUjfLoEr40bjbZaCxbIVS9fNZm9gPCjexcjJISFgInF34292CFtM4sK9 9WxdjFwcQgLbGSVWfN7LCOF8B3L+/YPKbGCU2HW5kw2khUVAVaLt7QIWEJsNyH7UDDFKWMBC YtvNf0wgNqeArcTDM/OYIFYoSPw59xisXkTASKL70ntWkKHMAhOYJA7famYESfAKeEucW3IS qIEDKGEm8W9ZCkRYUOLH5HtgvcwCOhK9378xQ9jiEs2tN6Hi2hJP3l1gBbEZBWQl3s2fzwoy RkTAUqLhkxzE2gKJ/3ffMkKcIyCxZM95ZghbVOLl43+sED8uZpG4eP8b0wRGiVkIV8xCcsUs JFfMQnLFLCRXLGBkXcUoVpxUlJmeUZKbmJmTbmCkl5Gpl5mXWrKJERJ3mTsYl+9UOcQowMGo xMOrsPVkoBBrYllxZe4hRgkOZiUR3jfzTwUK8aYkVlalFuXHF5XmpBYfYmTi4JRqYJwkuGVm +uuv7bwLuhZ53rjvZaLT//TIPJ013tJx30TZFN53iH3mm7Q0508b/5Yvsgd4BP/WT9DWeKr0 cNm6A9t+Njq+aop9Ns3w4pvn+T0XMlc310h4X2U81HrLe23rhmVbnq8/+y5/qUatqLa38W8T 7+K9rxdnOk7Y1NG5oCgvlcHjX0Tx1A9KLMUZiYZazEXFiQBhEme1mQIAAA== References: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> <6EB34CB5D82C4645B826C56144826EA97DF0AD29@szxeml560-mbx.china.huawei.com> Cc: "nsc@ietf.org" , "adrian@olddog.co.uk" , "Hongyu Li \(Julio\)" , " " , Ron Parker Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 14:38:25 -0000 Hi Jim, I agree that forwarding based on pushed headers is a possible model for bui= lding the chain, but others are possible as well. You can have a "service r= outer" where SFs (or NLFs) forward packets once processed and that decides = on the following step. Or you can build the chain by means of a SDN control= ler. Or it could even be conceivable to use some kind of specific addressin= g schema: a couple of possible ideas could be multicast or LISP... I do not= imply that I support any of these solutions (for the moment) or even any o= r all of them are necessarily feasible, but just to cite a few additional c= hoices at this point. Be goode, On 26 Jun 2013, at 16:16 , Jim Guichard (jguichar) wrote: > I think we need to be clear on definitions. I believe your comments to be= true (although I would not restrict it to the IP header only) for classifi= cation of traffic that needs to be forwarded into a service chain. Actual f= orwarding through the service chain is achieved using the headers pushed on= to the packets based upon the initial or subsequent classification. > > Sent from my iPhone > > On Jun 26, 2013, at 9:55 AM, "Diego R. Lopez" wrote: > >> Hi Ron, >> >> On 26 Jun 2013, at 15:13 , Ron Parker wrote: >>> >>> [snip] Another point I wanna make is that IMHO, on key character of NSC= is traffic is forwarded per service characteristics other than destination= addresses. This is the key value NSC WG, if will be formed, should work on= . Service characteristics may include the network service a use has subscri= bed to, e.g. parental control, firewall. That means a packet is forwarded i= nto a service chain, when a classifier matches some rules of service charac= teristics, which may be reflected by access ports, or segments or tuples in= a packet. >>> >>> [Ron]: I view NSC as a specific application of policy based forwarding= , where the policy is based on any or all of the fields in the IP header. = The end-to-end model of classifier-based nodes and non-classifier-based no= des means that there are 2 distinct sets of procedures to determine and app= ly the policy. >> >> >> I'd prefer to think that it is not necessarily limited to the fields in = the IP header. To start with, we could think in TCP/UDP headers or L2 ones.= Or even think about application stuff, like HTTP headers... >> >> Be goode, >> >> -- >> "Esta vez no fallaremos, Doctor Infierno" >> >> Dr Diego R. Lopez >> Telefonica I+D >> http://people.tid.es/diego.lopez/ >> >> e-mail: diego@tid.es >> Tel: +34 913 129 041 >> Mobile: +34 682 051 091 >> ----------------------------------------- >> >> >> ________________________________ >> >> Este mensaje se dirige exclusivamente a su destinatario. Puede consultar= nuestra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el = enlace situado m=E1s abajo. >> This message is intended exclusively for its addressee. We only send and= receive email on the basis of the terms set out at: >> http://www.tid.es/ES/PAGINAS/disclaimer.aspx -- "Esta vez no fallaremos, Doctor Infierno" Dr Diego R. Lopez Telefonica I+D http://people.tid.es/diego.lopez/ e-mail: diego@tid.es Tel: +34 913 129 041 Mobile: +34 682 051 091 ----------------------------------------- ________________________________ Este mensaje se dirige exclusivamente a su destinatario. Puede consultar nu= estra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el enl= ace situado m=E1s abajo. This message is intended exclusively for its addressee. We only send and re= ceive email on the basis of the terms set out at: http://www.tid.es/ES/PAGINAS/disclaimer.aspx From Ron_Parker@affirmednetworks.com Wed Jun 26 07:39:14 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id CDC5721E8121 for ; Wed, 26 Jun 2013 07:39:14 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.598 X-Spam-Level: X-Spam-Status: No, score=-2.598 tagged_above=-999 required=5 tests=[AWL=-0.000, BAYES_00=-2.599, HTML_MESSAGE=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id BJeRtOVm3rGy for ; Wed, 26 Jun 2013 07:39:08 -0700 (PDT) Received: from hub021-ca-7.exch021.serverdata.net (hub021-ca-7.exch021.serverdata.net [64.78.56.72]) by ietfa.amsl.com (Postfix) with ESMTP id C9FD421E8090 for ; Wed, 26 Jun 2013 07:39:08 -0700 (PDT) Received: from MBX021-W3-CA-2.exch021.domain.local ([10.254.4.78]) by HUB021-CA-7.exch021.domain.local ([10.254.4.109]) with mapi id 14.03.0123.003; Wed, 26 Jun 2013 07:39:08 -0700 From: Ron Parker To: "christian.jacquenet@orange.com" , "nsc@ietf.org" Thread-Topic: Comments on draft-boucadair-network-function-chaining-00 Thread-Index: Ac5x6/wxcUX+ReJaT8eNlB4kpr2L6QASIKPAABD+vdA= Date: Wed, 26 Jun 2013 14:39:07 +0000 Message-ID: References: <32603_1372228737_51CA8C81_32603_2651_1_983A1D8DA0DA5F4EB747BF34CBEE5CD15C418F0433@PUEXCB1C.nanterre.francetelecom.fr> In-Reply-To: <32603_1372228737_51CA8C81_32603_2651_1_983A1D8DA0DA5F4EB747BF34CBEE5CD15C418F0433@PUEXCB1C.nanterre.francetelecom.fr> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [50.203.66.100] Content-Type: multipart/alternative; boundary="_000_CDF2F015F4429F458815ED2A6C2B6B0B1A669A7BMBX021W3CA2exch_" MIME-Version: 1.0 Subject: Re: [nsc] Comments on draft-boucadair-network-function-chaining-00 X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 14:39:14 -0000 --_000_CDF2F015F4429F458815ED2A6C2B6B0B1A669A7BMBX021W3CA2exch_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Thanks for the response, Christian. My follow up is embedded. BR, Ron From: christian.jacquenet@orange.com [mailto:christian.jacquenet@orange.com= ] Sent: Wednesday, June 26, 2013 2:39 AM To: Ron Parker; nsc@ietf.org Subject: RE: Comments on draft-boucadair-network-function-chaining-00 Hello Ron, Many thanks for your feedback - please see inline. [snip] 1.3 Objectives Is it desirable to support unidirectional chains or are only bidirectional = chains considered? [CJ] I guess this depends on the nature of the service to be delivered and = the corresponding NLF functions to be invoked. Admittedly, the draft was or= iginally written with a kind of "request/response" type of service in mind,= but I think nothing really prevents unidirectional chaining scenarios - th= is could apply to a multicast service within the context of a QoS-based mul= ticast forwarding scheme that would solicit several functions in the access= infrastructure to make sure the multicast contents are properly delivered = to the receiver, for example. For bidirectional chains, is NLF chronology always reversed for each dir= ection, or is it arbitrarily controlled? [CJ] This should be arbitrarily controlled. Are NLF's identified by IP address, FQDN, or both? [CJ] I guess you refer to the NKF Identifier, as introduced in section 2, r= ight? As for the NLF Locator, we currently privilege the IP address, but FQ= DN could be used (see Section 3.2). As for the Identifier, this could be bo= th an IP address and a FQDN. Is it an objective for NLFC to explicitly handle load balancing towards mul= tiply-instantiated NLF's , or is load balancing in this regard considered t= o be part of each NLF? [CJ] I think it's primarily a design choice that will be conditioned by the= very nature of the NLF to be activated. I'm not sure each and every NLF we= can think about has implicit load balancing capabilities. Therefore, I wou= ld consider load balancing as a possible (but not necessarily mandatory) ob= jective of the NLFC machinery. [Ron] If the NLFC explicitly allows for load balancing by the PDP, then som= e requirements would follow wrt availability determination of the constitue= nt addressable entities. Optionally, a means to determine loading of the c= onstituent addressable entities so that the PDP may make a more informed de= cision. For the first, ping and BFD could be considered. For the second= , SNMP could be considered. Is it an objective for NLFC to determine the availability of an NLF? [CJ] No, I think the availability of an NLF suggests another mechanism, e.g= ., the (automated) discovery of the network topology, devices and their res= pective capabilities which will feed the PDP intelligence with information = to not only dynamically structure a network service but also to trigger NLF= C according to the outcomes of the negotiation of the service parameters be= tween the customer and the provider, for example. [Ron] End-to-end reliability and resiliency could be enhanced if the PDP co= uld prevent launching packets into a broken chain. Perhaps in some cases,= an unreachable NLF should be pruned from the chain while in other cases, t= he traffic should be dropped. A trap or other indication coming from the = PDP would assist in overall network visibility and troubleshooting. If so, are there policies regarding what to do when an NLF becomes unava= ilable? [CJ] To me, this is part of a feedback mechanism introduced in draft-sin-sd= nrg* which is meant to assess how efficiently a policy (or a set thereof) i= s enforced, but also to notify the PDP about any event (such as the unavail= ability of a NLF which failed the chaining operation for example) that may = affect the decision-making process. [Ron] My thinking here was a a proactive mechanism initiated by the PDP. 1.4 Assumptions All NLFs are reachable via standard IP routing techniques (i.e., no assumpt= ions regarding non-IP SDN protocols or techniques)? [CJ] Yes, the draft is generic enough to consider any IP networking environ= ment. 1.5 Rationale Suggest comment about fragmentation be moved to Objectives section. [CJ] Makes sense. 1.6 Generic Requirements Whether path stretch occurs or not is an administrative matter, assuming re= asonably correct IP routing topologies. [CJ] Agreed, but do you suggest any particular wording? [Ron] I think the requirement could be removed without diminishing the rest= of the specification. 3.3 Building NLF Maps Does an NLF map specify the set of NLF's, only, or does it specify the set = of NLF's and the policy to be applied by each. For example, assume that F= irewall-A and NAT64-A are distinct instances of NLFs. Is {Firewall-A-prof= ile-1, NAT64-A-profile-1} distinct from {Firewall-A-profile-2, Nat64-A-prof= ile-1}. Or, alternatively, is the policy conveyed outside of the map? [CJ] Policies are conveyed outside of the map. [Ron] My thinking here is that some NLF's are capable of providing differen= tiated services (i.e., a DPI function could block peer-to-peer for one grou= p of users and voip for another group of users). If the DPI function is a= non-classification node (in NLFC parlance), then it may be beneficial for = the PDP to indicate which differentiated service to apply. One approach i= s that the NLFC map explicitly indicates this. If that approach were foll= owed, then an NLF's containment within a particular NLF Map would indicate = 2 things - first, that the NLF should be a member of the chain with a parti= cular previous hop and a particular next hop (this is pretty obvious), and = second, that the NLF should apply a particular local policy (where coordina= tion of the NLFC map index and the local policy is an administrative matter= ). 4.1 NLFC Boundary Node I think the definition of ingress/egress here is problematic as it seems to= be implying topological direction (i.e., uplink and downlink in a mobile e= nvironment). If, however, the NLFC domain is thought of as residing in be= tween 2 non-NLFC domains (or segments of the same domain), then there is bo= th ingress and egress at both ends of the chain. A more descriptive appro= ach may be to use terms such as "West-ingress", "East-egress", "East-ingres= s", "West-egress". In a wireless domain, perhaps "north" and "south" wou= ld be more intuitive, but I think the connotation is more limiting in a gen= eral sense. [CJ] I would avoid that kind of elaboration. As for the case you rightfully= mention, I think we can still deal with ingress and egress notions on both= "sides" of the NLFC domain. [Ron] I may have read more into the original language than was intended. 4.3 NLFC Ingress Node As an ingress node which connects to the non NLFC domain, when would an NLF= C Map Index be present? [CJ] I guess the operation is similar to DSCP-re-marking, with implicitly s= ecurity threats that have been mentioned in the document. Also, this implies an encapsulation technique that has not been previousl= y introduced in the document. 5.1.1 NLFC Map Index If the individual entries in an NLFC Map indicate only the identity of the = NLF, then 8 or 16 bits to identify which NLFC Map would be sufficient. If= , however, the entries in an NLFC Map indicate NLF identity+policy, then 16= or 32 bits would be more appropriate. [CJ] Right. [Ron] See discussion of local policy above. 5.1.3 Where to store NLFC Map Index in a packet? Consider usage of GRE key. Advantage is that in many cases, encapsulation= is required and GRE is a standard encapsulation technique which provides a= 32-bit general purpose field. Disadvantage is that some systems may requ= ire each unique GRE key to be provisioned as an IP interface. [CJ] OK. Thanks, Christian. ___________________________________________________________________________= ______________________________________________ Ce message et ses pieces jointes peuvent contenir des informations confiden= tielles ou privilegiees et ne doivent donc pas etre diffuses, exploites ou copies sans autorisation. Si vous avez recu= ce message par erreur, veuillez le signaler a l'expediteur et le detruire ainsi que les pieces jointes. Les messages el= ectroniques etant susceptibles d'alteration, Orange decline toute responsabilite si ce message a ete altere, deforme ou = falsifie. Merci. This message and its attachments may contain confidential or privileged inf= ormation that may be protected by law; they should not be distributed, used or copied without authorisation. If you have received this email in error, please notify the sender and dele= te this message and its attachments. As emails may be altered, Orange is not liable for messages that have been = modified, changed or falsified. Thank you. --_000_CDF2F015F4429F458815ED2A6C2B6B0B1A669A7BMBX021W3CA2exch_ Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable

Thanks for the respons= e, Christian.

 

My follow up is embedd= ed.

 

BR,<= /p>

Ron<= /p>

 

 

From: christian.jacquenet@orange.com [mailto:= christian.jacquenet@orange.com]
Sent: Wednesday, June 26, 2013 2:39 AM
To: Ron Parker; nsc@ietf.org
Subject: RE: Comments on draft-boucadair-network-function-chaining-0= 0

 

Hello Ron,

 

Many thanks for your feedback – please= see inline.

 

[snip]

 

1.3 Objectives=

Is it desirable to suppor= t unidirectional chains or are only bidirectional chains considered?

[CJ] I guess this depends on the nature of t= he service to be delivered and the corresponding NLF functions to be invoke= d. Admittedly, the draft was originally written with a kind of “request/response” type of service in mind, but= I think nothing really prevents unidirectional chaining scenarios – = this could apply to a multicast service within the context of a QoS-based m= ulticast forwarding scheme that would solicit several functions in the access infrastructure to make sure the multicast contents= are properly delivered to the receiver, for example.

   For bidirect= ional chains, is NLF chronology always reversed for each direction, or is i= t arbitrarily  controlled?

[CJ] This should be arbitrarily controlled.<= o:p>

Are NLF’s identifie= d by IP address, FQDN, or both?

[CJ] I guess you refer to the NKF Identifier= , as introduced in section 2, right? As for the NLF Locator, we currently p= rivilege the IP address, but FQDN could be used (see Section 3.2). As for the Identifier, this could be both an IP address= and a FQDN.

Is it an objective for NL= FC to explicitly handle load balancing towards multiply-instantiated NLF= 217;s , or is load balancing in this regard considered to be part of each N= LF?

[CJ] I think it’s primarily a design c= hoice that will be conditioned by the very nature of the NLF to be activate= d. I’m not sure each and every NLF we can think about has implicit load balancing capabilities. Therefore, I would consider load= balancing as a possible (but not necessarily mandatory) objective of the N= LFC machinery.

[Ron] If the NLFC expl= icitly allows for load balancing by the PDP, then some requirements would f= ollow wrt availability determination of the constituent addressable entitie= s.  Optionally, a means to determine loading of the constituent addressable entities so that the PDP may make a= more informed decision.   For the first, ping and BFD could be c= onsidered.   For the second, SNMP could be considered.=

 

Is it an objective for NL= FC to determine the availability of an NLF?

[CJ] No, I think the availability of an NLF = suggests another mechanism, e.g., the (automated) discovery of the network = topology, devices and their respective capabilities which will feed the PDP intelligence with information to not only dynamica= lly structure a network service but also to trigger NLFC according to the o= utcomes of the negotiation of the service parameters between the customer a= nd the provider, for example.

[Ron] End-to-end relia= bility and resiliency could be enhanced if the PDP could prevent launching = packets into a broken chain.   Perhaps in some cases, an unreacha= ble NLF should be pruned from the chain while in other cases, the traffic should be dropped.   A trap or other= indication coming from the PDP would assist in overall network visibility = and troubleshooting.

 

   If so, are t= here policies regarding what to do when an NLF becomes unavailable?

[CJ] To me, this is part of a feedback mecha= nism introduced in draft-sin-sdnrg* which is meant to assess how efficientl= y a policy (or a set thereof) is enforced, but also to notify the PDP about any event (such as the unavailability of a NL= F which failed the chaining operation for example) that may affect the deci= sion-making process.

[Ron]  My thinking here was a a proacti= ve mechanism initiated by the PDP.

 

1.4 Assumptions

All NLFs are reachable vi= a standard IP routing techniques (i.e., no assumptions regarding non-IP SDN= protocols or techniques)?

[CJ] Yes, the draft is generic enough to con= sider any IP networking environment.

 

1.5 Rationale<= /p>

Suggest comment about fra= gmentation be moved to Objectives section.

[CJ] Makes sense.

 

1.6 Generic Requirements<= o:p>

Whether path stretch occu= rs or not is an administrative matter, assuming reasonably correct IP routi= ng topologies.

[CJ] Agreed, but do you suggest any particul= ar wording?

[Ron] I think the requirement could be remov= ed without diminishing the rest of the specification.

 

3.3 Building NLF Maps

Does an NLF map specify t= he set of NLF’s, only, or does it specify the set of NLF’s and = the policy to be applied by each.   For example, assume that Fire= wall-A and NAT64-A are distinct instances of NLFs.   Is {Firewall= -A-profile-1, NAT64-A-profile-1} distinct from {Firewall-A-profile-2, Nat64-A-profile-1}= .   Or, alternatively, is the policy conveyed outside of the map?=

[CJ] Policies are conveyed outside of the ma= p.

[Ron] My thinking here is that some NLF̵= 7;s are capable of providing differentiated services (i.e., a DPI function = could block peer-to-peer for one group of users and voip for another group of users).   If the DPI function is a non= -classification node (in NLFC parlance), then it may be beneficial for the = PDP to indicate which differentiated service to apply.   One appr= oach is that the NLFC map explicitly indicates this.   If that approach were followed, then an NLF’s containment within a p= articular NLF Map would indicate 2 things – first, that the NLF shoul= d be a member of the chain with a particular previous hop and a particular = next hop (this is pretty obvious), and second, that the NLF should apply a particular local policy (where coordination of= the NLFC map index and the local policy is an administrative matter). = ;

 

4.1  NLFC Boundary N= ode

I think the definition of= ingress/egress here is problematic as it seems to be implying topological = direction (i.e., uplink and downlink in a mobile environment).   = If, however, the NLFC domain is thought of as residing in between 2 non-NLFC domains (or segments of the same domain), t= hen there is both ingress and egress at both ends of the chain.  = A more descriptive approach may be to use terms such as “West-ingres= s”, “East-egress”, “East-ingress”, “Wes= t-egress”.    In a wireless domain, perhaps “north” and “south” = would be more intuitive, but I think the connotation is more limiting in a = general sense.

[CJ] I would avoid that kind of elaboration.= As for the case you rightfully mention, I think we can still deal with ing= ress and egress notions on both “sides” of the NLFC domain.

[Ron] I may have read more into the original= language than was intended.

 

4.3 NLFC Ingress Node

As an ingress node which = connects to the non NLFC domain, when would an NLFC Map Index be present?&n= bsp;

[CJ] I guess the operation is similar to DSC= P-re-marking, with implicitly security threats that have been mentioned in = the document.

  Also, this implies= an encapsulation technique that has not been previously introduced in the = document.

 

5.1.1 NLFC Map Index=

If the individual entries= in an NLFC Map indicate only the identity of the NLF, then 8 or 16 bits to= identify which NLFC Map would be sufficient.   If, however, the = entries in an NLFC Map indicate NLF identity+policy, then 16 or 32 bits would be more appropriate.

[CJ] Right.

[Ron] See discussion of local policy above.<= o:p>

 

5.1.3 Where to store NLFC= Map Index in a packet?

Consider usage of GRE key= .   Advantage is that in many cases, encapsulation is required an= d GRE is a standard encapsulation technique which provides a 32-bit general= purpose field.   Disadvantage is that some systems may require each unique GRE key to be provisioned as an IP interface.=

[CJ] OK.

 

Thanks,

 

Christian.

 

______________________________________________________________________=
___________________________________________________
 
Ce message et ses pieces jointes peuvent contenir des informations con=
fidentielles ou privilegiees et ne doivent donc
pas etre diffuses, exploites ou copies sans autorisation. Si vous avez=
 recu ce message par erreur, veuillez le signaler
a l'expediteur et le detruire ainsi que les pieces jointes. Les messag=
es electroniques etant susceptibles d'alteration,
Orange decline toute responsabilite si ce message a ete altere, deform=
e ou falsifie. Merci.
 
This message and its attachments may contain confidential or privilege=
d information that may be protected by law;
they should not be distributed, used or copied without authorisation.<=
o:p>
If you have received this email in error, please notify the sender and=
 delete this message and its attachments.
As emails may be altered, Orange is not liable for messages that have =
been modified, changed or falsified.
Thank you.
--_000_CDF2F015F4429F458815ED2A6C2B6B0B1A669A7BMBX021W3CA2exch_-- From jguichar@cisco.com Wed Jun 26 07:45:13 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9D95C11E8116 for ; Wed, 26 Jun 2013 07:45:13 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.599 X-Spam-Level: X-Spam-Status: No, score=-10.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id zZTc1jjq4A+5 for ; Wed, 26 Jun 2013 07:45:08 -0700 (PDT) Received: from rcdn-iport-7.cisco.com (rcdn-iport-7.cisco.com [173.37.86.78]) by ietfa.amsl.com (Postfix) with ESMTP id CEE5F21F9989 for ; Wed, 26 Jun 2013 07:45:07 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=4107; q=dns/txt; s=iport; t=1372257908; x=1373467508; h=from:to:cc:subject:date:message-id:references: in-reply-to:content-transfer-encoding:mime-version; bh=0sxXsgdUyRix8McKP2uftxEFLe3jjO91pbr2EeRh+54=; b=GaK/IHWsVp7m/oF4xMAPdaY3QNqeQZqlF4H89OUfQEAYp4tRvuvqX1OE p5mJ+N1xZLgmIR79FMwnXRfxKL0+zfGAsRbYjHAcvC3RxYyGUcRQY8/xp ZlnSate32trbZDvhStz/JsYXhy11Ez9kDJKh1tvyumanDuRg595oL4dVz k=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AhAFAFj9ylGtJV2a/2dsb2JhbABagwkxvjqBCoEAFnSCIwEBAQMBeQULAgEIGC4yJQIEDgUbh20GDLphjgkKBn8zB4MCYQOIao5bgSmJBocWgxGBaAkX X-IronPort-AV: E=Sophos;i="4.87,944,1363132800"; d="scan'208";a="227666884" Received: from rcdn-core-3.cisco.com ([173.37.93.154]) by rcdn-iport-7.cisco.com with ESMTP; 26 Jun 2013 14:45:07 +0000 Received: from xhc-aln-x11.cisco.com (xhc-aln-x11.cisco.com [173.36.12.85]) by rcdn-core-3.cisco.com (8.14.5/8.14.5) with ESMTP id r5QEj7LF029689 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Wed, 26 Jun 2013 14:45:07 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.80]) by xhc-aln-x11.cisco.com ([173.36.12.85]) with mapi id 14.02.0318.004; Wed, 26 Jun 2013 09:45:07 -0500 From: "Jim Guichard (jguichar)" To: "Diego R. Lopez" Thread-Topic: [nsc] What is a service? Why is NSC not SC? Thread-Index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewABC13ID//+4ggP//KpNAgAYtNICAA2WvAIAAremAgAALUgD//7KLOIAAWX4A//+uY8g= Date: Wed, 26 Jun 2013 14:45:05 +0000 Message-ID: <3CC2C329-36CA-4628-BAD1-5DDD21D2C1B0@cisco.com> References: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> <6EB34CB5D82C4645B826C56144826EA97DF0AD29@szxeml560-mbx.china.huawei.com> , In-Reply-To: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Cc: "nsc@ietf.org" , "adrian@olddog.co.uk" , "Hongyu Li \(Julio\)" , " " , Ron Parker Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 14:45:13 -0000 Yes I agree Diego but my point was that classification may be a different o= peration to forwarding. Sent from my iPhone On Jun 26, 2013, at 10:38 AM, "Diego R. Lopez" wrote: > Hi Jim, >=20 > I agree that forwarding based on pushed headers is a possible model for b= uilding the chain, but others are possible as well. You can have a "service= router" where SFs (or NLFs) forward packets once processed and that decide= s on the following step. Or you can build the chain by means of a SDN contr= oller. Or it could even be conceivable to use some kind of specific address= ing schema: a couple of possible ideas could be multicast or LISP... I do n= ot imply that I support any of these solutions (for the moment) or even any= or all of them are necessarily feasible, but just to cite a few additional= choices at this point. >=20 > Be goode, >=20 > On 26 Jun 2013, at 16:16 , Jim Guichard (jguichar) wrote: >=20 >> I think we need to be clear on definitions. I believe your comments to b= e true (although I would not restrict it to the IP header only) for classif= ication of traffic that needs to be forwarded into a service chain. Actual = forwarding through the service chain is achieved using the headers pushed o= n to the packets based upon the initial or subsequent classification. >>=20 >> Sent from my iPhone >>=20 >> On Jun 26, 2013, at 9:55 AM, "Diego R. Lopez" wrote: >>=20 >>> Hi Ron, >>>=20 >>> On 26 Jun 2013, at 15:13 , Ron Parker wrote: >>>>=20 >>>> [snip] Another point I wanna make is that IMHO, on key character of NS= C is traffic is forwarded per service characteristics other than destinatio= n addresses. This is the key value NSC WG, if will be formed, should work o= n. Service characteristics may include the network service a use has subscr= ibed to, e.g. parental control, firewall. That means a packet is forwarded = into a service chain, when a classifier matches some rules of service chara= cteristics, which may be reflected by access ports, or segments or tuples i= n a packet. >>>>=20 >>>> [Ron]: I view NSC as a specific application of policy based forwardin= g, where the policy is based on any or all of the fields in the IP header. = The end-to-end model of classifier-based nodes and non-classifier-based n= odes means that there are 2 distinct sets of procedures to determine and ap= ply the policy. >>>=20 >>>=20 >>> I'd prefer to think that it is not necessarily limited to the fields in= the IP header. To start with, we could think in TCP/UDP headers or L2 ones= . Or even think about application stuff, like HTTP headers... >>>=20 >>> Be goode, >>>=20 >>> -- >>> "Esta vez no fallaremos, Doctor Infierno" >>>=20 >>> Dr Diego R. Lopez >>> Telefonica I+D >>> http://people.tid.es/diego.lopez/ >>>=20 >>> e-mail: diego@tid.es >>> Tel: +34 913 129 041 >>> Mobile: +34 682 051 091 >>> ----------------------------------------- >>>=20 >>>=20 >>> ________________________________ >>>=20 >>> Este mensaje se dirige exclusivamente a su destinatario. Puede consulta= r nuestra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el= enlace situado m=E1s abajo. >>> This message is intended exclusively for its addressee. We only send an= d receive email on the basis of the terms set out at: >>> http://www.tid.es/ES/PAGINAS/disclaimer.aspx >=20 >=20 > -- > "Esta vez no fallaremos, Doctor Infierno" >=20 > Dr Diego R. Lopez > Telefonica I+D > http://people.tid.es/diego.lopez/ >=20 > e-mail: diego@tid.es > Tel: +34 913 129 041 > Mobile: +34 682 051 091 > ----------------------------------------- >=20 >=20 > ________________________________ >=20 > Este mensaje se dirige exclusivamente a su destinatario. Puede consultar = nuestra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el e= nlace situado m=E1s abajo. > This message is intended exclusively for its addressee. We only send and = receive email on the basis of the terms set out at: > http://www.tid.es/ES/PAGINAS/disclaimer.aspx From diego@tid.es Wed Jun 26 07:45:19 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D7B7C21E809A for ; Wed, 26 Jun 2013 07:45:19 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 8+Mbt72P4oHd for ; Wed, 26 Jun 2013 07:45:15 -0700 (PDT) Received: from correo-bck.tid.es (correo-bck.tid.es [195.235.93.200]) by ietfa.amsl.com (Postfix) with ESMTP id D7A2B11E8116 for ; Wed, 26 Jun 2013 07:45:14 -0700 (PDT) Received: from sbrightmailg02.hi.inet (Sbrightmailg02.hi.inet [10.95.78.105]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTP id <0MP000JAN8ZEUW@tid.hi.inet> for nsc@ietf.org; Wed, 26 Jun 2013 16:45:14 +0200 (MEST) Received: from vanvan (vanvan.hi.inet [10.95.78.49]) by sbrightmailg02.hi.inet (Symantec Messaging Gateway) with SMTP id 37.0A.02911.97EFAC15; Wed, 26 Jun 2013 16:45:13 +0200 (CEST) Received: from correo.tid.es (mailhost.hi.inet [10.95.64.100]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTPS id <0MP000JAB8ZDUW@tid.hi.inet> for nsc@ietf.org; Wed, 26 Jun 2013 16:45:13 +0200 (MEST) Received: from EX10-MB2-MAD.hi.inet ([169.254.2.38]) by EX10-HTCAS8-MAD.hi.inet ([fe80::41c8:e965:8a6:de67%11]) with mapi id 14.02.0328.009; Wed, 26 Jun 2013 16:45:13 +0200 Date: Wed, 26 Jun 2013 14:44:09 +0000 From: "Diego R. Lopez" In-reply-to: X-Originating-IP: [10.95.64.115] To: Ron Parker Message-id: Content-id: <4D3B4E5EBBABB7438756D8E59C8C87C3@hi.inet> MIME-version: 1.0 Content-type: text/plain; charset=iso-8859-1 Content-language: en-US Content-transfer-encoding: quoted-printable Accept-Language: en-US, es-ES Thread-topic: [nsc] What is a service? Why is NSC not SC? Thread-index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewAAIKxYAABiYCAAAeHsEAAISBAQAAbK0RAAAVvSOAAAFzfAAAACKSAAABnJ2A X-AuditID: 0a5f4e69-b7f118e000000b5f-e6-51cafe794e98 X-MS-Has-Attach: X-MS-TNEF-Correlator: X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFmpnkeLIzCtJLcpLzFFi42Lhivcz1K38dyrQ4E63tcXxffvZHBg9liz5 yRTAGMVlk5Kak1mWWqRvl8CVcaLvLlvBErmKnQsmsDUwbpHoYuTkkBAwkfjYspAFwhaTuHBv PVsXIxeHkMB2RomJE1uhnO+MEjf//2KGcDYwSvz72w3WwiKgKrH41E4mEJsNyH7U/JsdxBYW sJDYdvMfUJyDg1MgVqJ9ZjzEBgWJP+ceg7WKCBhIvP58ihVkJrPAJCaJPQvusIEkeAW8Jfou XAGzmQXMJDbubmKCiAtK/Jh8jwUiriPR+/0bM4QtLtHcehMqri3x5N0FVhCbUUBW4t38+awg N4gIWEo0fJKD2Fsg0XvmHCPEPQISS/acZ4awRSVePv7HCvHjQRaJls37WScwSsxCcsYsJGfM QnLGLCRnzEJyxgJG1lWMYsVJRZnpGSW5iZk56QZGehmZepl5qSWbGCGRl7mDcflOlUOMAhyM Sjy8CltPBgqxJpYVV+YeYpTgYFYS4X0z/1SgEG9KYmVValF+fFFpTmrxIUYmDk6pBsZjv24p cV6du8Z18Y1939Z1fPP61TQ/6/acRVsjfhs23VizcrN7jLe/PJ8Cs/+acy2zfl3m5upmqbhm JHjn/1JXc+nvaX9SWUvCmXv+lk0WeLli0g0eyzSTz3OXWMdV//3zkX3yzmuSG++8jX67oz5k 28QVp6q5T6asVnZi42ir7X7TJdmSt9ZeiaU4I9FQi7moOBEAtfUoOJoCAAA= References: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> <6EB34CB5D82C4645B826C56144826EA97DF0AD29@szxeml560-mbx.china.huawei.com> Cc: "nsc@ietf.org" , "Jim Guichard \(jguichar\)" , "Hongyu Li \(Julio\)" , " " , "adrian@olddog.co.uk" Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 14:45:20 -0000 Hi Ron, Agreed. That's why a few messages ago I was saying we needed to define how = to identify the SF/NLF interfaces... Be goode, On 26 Jun 2013, at 15:59 , Ron Parker wrote: > Hi, Diego. > > Thanks for the response. Comment below. > > BR, > Ron > > > [snip] I'd prefer to think that it is not necessarily limited to the fiel= ds in the IP header. To start with, we could think in TCP/UDP headers or L2= ones. Or even think about application stuff, like HTTP headers... > > [Ron] Agree completely. NSC is a specific application of policy based f= orwarding where the policy is based on classification of the packets, inclu= ding layers 2 through 7. By stating that the NLF's are located via thei= r IP addresses and/or FQDN's, we don't require or preclude any particular f= orwarding paradigm between the classifier node and the non-classifier node. > > > -----Original Message----- > From: Diego R. Lopez [mailto:diego@tid.es] > Sent: Wednesday, June 26, 2013 9:54 AM > To: Ron Parker > Cc: Hongyu Li (Julio); ; nsc@ietf.org; Jim Guichard (jguichar); adrian@olddog.co.uk > Subject: Re: [nsc] What is a service? Why is NSC not SC? > > Hi Ron, > > On 26 Jun 2013, at 15:13 , Ron Parker wrote: >> >> [snip] Another point I wanna make is that IMHO, on key character of NSC = is traffic is forwarded per service characteristics other than destination = addresses. This is the key value NSC WG, if will be formed, should work on.= Service characteristics may include the network service a use has subscrib= ed to, e.g. parental control, firewall. That means a packet is forwarded in= to a service chain, when a classifier matches some rules of service charact= eristics, which may be reflected by access ports, or segments or tuples in = a packet. >> >> [Ron]: I view NSC as a specific application of policy based forwarding,= where the policy is based on any or all of the fields in the IP header. = The end-to-end model of classifier-based nodes and non-classifier-based nod= es means that there are 2 distinct sets of procedures to determine and appl= y the policy. > > > I'd prefer to think that it is not necessarily limited to the fields in t= he IP header. To start with, we could think in TCP/UDP headers or L2 ones. = Or even think about application stuff, like HTTP headers... > > Be goode, > > -- > "Esta vez no fallaremos, Doctor Infierno" > > Dr Diego R. Lopez > Telefonica I+D > http://people.tid.es/diego.lopez/ > > e-mail: diego@tid.es > Tel: +34 913 129 041 > Mobile: +34 682 051 091 > ----------------------------------------- > > > ________________________________ > > Este mensaje se dirige exclusivamente a su destinatario. Puede consultar = nuestra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el e= nlace situado m=E1s abajo. > This message is intended exclusively for its addressee. We only send and = receive email on the basis of the terms set out at: > http://www.tid.es/ES/PAGINAS/disclaimer.aspx > _______________________________________________ > nsc mailing list > nsc@ietf.org > https://www.ietf.org/mailman/listinfo/nsc -- "Esta vez no fallaremos, Doctor Infierno" Dr Diego R. Lopez Telefonica I+D http://people.tid.es/diego.lopez/ e-mail: diego@tid.es Tel: +34 913 129 041 Mobile: +34 682 051 091 ----------------------------------------- ________________________________ Este mensaje se dirige exclusivamente a su destinatario. Puede consultar nu= estra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el enl= ace situado m=E1s abajo. This message is intended exclusively for its addressee. We only send and re= ceive email on the basis of the terms set out at: http://www.tid.es/ES/PAGINAS/disclaimer.aspx From diego@tid.es Wed Jun 26 07:46:43 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 88D3B21F9057 for ; Wed, 26 Jun 2013 07:46:43 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Vi4uijz-EbwY for ; Wed, 26 Jun 2013 07:46:39 -0700 (PDT) Received: from tidos.tid.es (tidos.tid.es [195.235.93.44]) by ietfa.amsl.com (Postfix) with ESMTP id 5177D21F9FB3 for ; Wed, 26 Jun 2013 07:46:38 -0700 (PDT) Received: from sbrightmailg01.hi.inet (sbrightmailg01.hi.inet [10.95.64.104]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTP id <0MP0003NJ91OCD@tid.hi.inet> for nsc@ietf.org; Wed, 26 Jun 2013 16:46:37 +0200 (MEST) Received: from tid (tid.hi.inet [10.95.64.10]) by sbrightmailg01.hi.inet (Symantec Messaging Gateway) with SMTP id 7C.D9.03142.CCEFAC15; Wed, 26 Jun 2013 16:46:37 +0200 (CEST) Received: from correo.tid.es (mailhost.hi.inet [10.95.64.100]) by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8 2006)) with ESMTP id <0MP0003ND91OCD@tid.hi.inet> for nsc@ietf.org; Wed, 26 Jun 2013 16:46:36 +0200 (MEST) Received: from EX10-MB2-MAD.hi.inet ([169.254.2.38]) by EX10-HTCAS6-MAD.hi.inet ([::1]) with mapi id 14.02.0328.009; Wed, 26 Jun 2013 16:46:36 +0200 Date: Wed, 26 Jun 2013 14:45:33 +0000 From: "Diego R. Lopez" In-reply-to: <3CC2C329-36CA-4628-BAD1-5DDD21D2C1B0@cisco.com> X-Originating-IP: [10.95.64.115] To: "Jim Guichard (jguichar)" Message-id: Content-id: <8E84B151411740479802D8596C480D20@hi.inet> MIME-version: 1.0 Content-type: text/plain; charset=iso-8859-1 Content-language: en-US Content-transfer-encoding: quoted-printable Accept-Language: en-US, es-ES Thread-topic: [nsc] What is a service? Why is NSC not SC? Thread-index: Ac5tqI54HknBNelNRGizJvx+66ywewAALrewAAIKxYAABiYCAAAeHsEAAISBAQAAbK0RAAAVvSOAAAFzfAAAAMIqgAAAvuIAAAA9PoAAAA1qgA== X-AuditID: 0a5f4068-b7f128e000000c46-46-51cafecc33e1 X-MS-Has-Attach: X-MS-TNEF-Correlator: X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFupkkeLIzCtJLcpLzFFi42Lhinfg0j3771SgwdKtTBbH9+1nc2D0WLLk J1MAYxSXTUpqTmZZapG+XQJXRuO0KSwFz1Urziw9wN7AuFuui5GTQ0LARKL76kUmCFtM4sK9 9WxdjFwcQgIbGSXWvnrMDOF8Y5TYv2IulDONUeLA3g3MIC0sAqoSB/a1sILYbED2o+bf7CC2 sICFxLab/8DGcgrYSlzd8YMFYoWCxJ9zj8FsEQEjie5L71lBhjILTGCSOHyrmREkwSvgLTHz KMQCZgEziYVfTrFAxAUlfky+xwIR15Ho/f4NqkZcorn1JlRcW+LJuwtgBzEKyEq8mz8fyOYA WmYp0fBJDmSXiEADo8SVOTOhDhKQWLLnPDOELSrx8vE/VogvZ7BKHFz6hHECo8QsJHfMQnLH LCR3zEJyxywkdyxgZF3FKFacVJSZnlGSm5iZk25gqJeRqZeZl1qyiRESfRk7GJfvVDnEKMDB qMTDe2LTyUAh1sSy4srcQ4wSHMxKIrxv5p8KFOJNSaysSi3Kjy8qzUktPsTIxMEp1cDowfhk lsickvXmNl/8674aJzI8DPFo5j425brMhgnH7tckvhVfYVm4+VMUn3/CySUV7jJz4+yfR3pH WK14E7O5xsm32Ka9KD1oyePg5dtiDr7p1+qYytpSG9+uNuXctoCeDf0s+XlXa95cdL3x7ILp rPlNCa8Lv87UL3bKW+XacKB99pL3YRuUWIozEg21mIuKEwHxJMRrnAIAAA== References: <080401ce6dc2$388d2420$a9a76c60$@olddog.co.uk> <68B171751455884590F8E38E96416F363F53FEB3@xmb-rcd-x01.cisco.com> <94C682931C08B048B7A8645303FDC9F36EDB25954C@PUEXCB1B.nanterre.francetelecom.fr> <6EB34CB5D82C4645B826C56144826EA97DF0AD29@szxeml560-mbx.china.huawei.com> <3CC2C329-36CA-4628-BAD1-5DDD21D2C1B0@cisco.com> Cc: "nsc@ietf.org" , "adrian@olddog.co.uk" , "Hongyu Li \(Julio\)" , " " , Ron Parker Subject: Re: [nsc] What is a service? Why is NSC not SC? X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 14:46:43 -0000 And I can only say "amen" to that! On 26 Jun 2013, at 16:45 , Jim Guichard (jguichar) wrote: > Yes I agree Diego but my point was that classification may be a different= operation to forwarding. > > Sent from my iPhone > > On Jun 26, 2013, at 10:38 AM, "Diego R. Lopez" wrote: > >> Hi Jim, >> >> I agree that forwarding based on pushed headers is a possible model for = building the chain, but others are possible as well. You can have a "servic= e router" where SFs (or NLFs) forward packets once processed and that decid= es on the following step. Or you can build the chain by means of a SDN cont= roller. Or it could even be conceivable to use some kind of specific addres= sing schema: a couple of possible ideas could be multicast or LISP... I do = not imply that I support any of these solutions (for the moment) or even an= y or all of them are necessarily feasible, but just to cite a few additiona= l choices at this point. >> >> Be goode, >> >> On 26 Jun 2013, at 16:16 , Jim Guichard (jguichar) wrote: >> >>> I think we need to be clear on definitions. I believe your comments to = be true (although I would not restrict it to the IP header only) for classi= fication of traffic that needs to be forwarded into a service chain. Actual= forwarding through the service chain is achieved using the headers pushed = on to the packets based upon the initial or subsequent classification. >>> >>> Sent from my iPhone >>> >>> On Jun 26, 2013, at 9:55 AM, "Diego R. Lopez" wrote: >>> >>>> Hi Ron, >>>> >>>> On 26 Jun 2013, at 15:13 , Ron Parker wrote: >>>>> >>>>> [snip] Another point I wanna make is that IMHO, on key character of N= SC is traffic is forwarded per service characteristics other than destinati= on addresses. This is the key value NSC WG, if will be formed, should work = on. Service characteristics may include the network service a use has subsc= ribed to, e.g. parental control, firewall. That means a packet is forwarded= into a service chain, when a classifier matches some rules of service char= acteristics, which may be reflected by access ports, or segments or tuples = in a packet. >>>>> >>>>> [Ron]: I view NSC as a specific application of policy based forwardi= ng, where the policy is based on any or all of the fields in the IP header.= The end-to-end model of classifier-based nodes and non-classifier-based = nodes means that there are 2 distinct sets of procedures to determine and a= pply the policy. >>>> >>>> >>>> I'd prefer to think that it is not necessarily limited to the fields i= n the IP header. To start with, we could think in TCP/UDP headers or L2 one= s. Or even think about application stuff, like HTTP headers... >>>> >>>> Be goode, >>>> >>>> -- >>>> "Esta vez no fallaremos, Doctor Infierno" >>>> >>>> Dr Diego R. Lopez >>>> Telefonica I+D >>>> http://people.tid.es/diego.lopez/ >>>> >>>> e-mail: diego@tid.es >>>> Tel: +34 913 129 041 >>>> Mobile: +34 682 051 091 >>>> ----------------------------------------- >>>> >>>> >>>> ________________________________ >>>> >>>> Este mensaje se dirige exclusivamente a su destinatario. Puede consult= ar nuestra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en e= l enlace situado m=E1s abajo. >>>> This message is intended exclusively for its addressee. We only send a= nd receive email on the basis of the terms set out at: >>>> http://www.tid.es/ES/PAGINAS/disclaimer.aspx >> >> >> -- >> "Esta vez no fallaremos, Doctor Infierno" >> >> Dr Diego R. Lopez >> Telefonica I+D >> http://people.tid.es/diego.lopez/ >> >> e-mail: diego@tid.es >> Tel: +34 913 129 041 >> Mobile: +34 682 051 091 >> ----------------------------------------- >> >> >> ________________________________ >> >> Este mensaje se dirige exclusivamente a su destinatario. Puede consultar= nuestra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el = enlace situado m=E1s abajo. >> This message is intended exclusively for its addressee. We only send and= receive email on the basis of the terms set out at: >> http://www.tid.es/ES/PAGINAS/disclaimer.aspx -- "Esta vez no fallaremos, Doctor Infierno" Dr Diego R. Lopez Telefonica I+D http://people.tid.es/diego.lopez/ e-mail: diego@tid.es Tel: +34 913 129 041 Mobile: +34 682 051 091 ----------------------------------------- ________________________________ Este mensaje se dirige exclusivamente a su destinatario. Puede consultar nu= estra pol=EDtica de env=EDo y recepci=F3n de correo electr=F3nico en el enl= ace situado m=E1s abajo. This message is intended exclusively for its addressee. We only send and re= ceive email on the basis of the terms set out at: http://www.tid.es/ES/PAGINAS/disclaimer.aspx From paulq@cisco.com Wed Jun 26 08:32:56 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 4AB7611E811B for ; Wed, 26 Jun 2013 08:32:56 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.599 X-Spam-Level: X-Spam-Status: No, score=-10.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 1Ox2+njsQk1B for ; Wed, 26 Jun 2013 08:32:51 -0700 (PDT) Received: from mtv-iport-3.cisco.com (mtv-iport-3.cisco.com [173.36.130.14]) by ietfa.amsl.com (Postfix) with ESMTP id E59EB11E81C5 for ; Wed, 26 Jun 2013 08:32:48 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=8401; q=dns/txt; s=iport; t=1372260769; x=1373470369; h=mime-version:subject:from:in-reply-to:date:cc: content-transfer-encoding:message-id:references:to; bh=oNuz2bQ0GBQ2pMQakw2nX0vcYPk+qgepgxcmmWDEQgs=; b=gHXPAsLuGzhUvFT07SwB/nBGuUNga44tPRskYS7Ct+4+aAIUaUM5G5xt 5HWc3y6qf6omZGnksOEMRQSXORMGhHJODW38Er9SmGs0vF+qb/ZXFeIiK Xz0FjYRaaU/yIkXAyG3UyDuyA8ltiK6go1wq5e1YcNxO8NN2+x8hJwUWA A=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AgIFABIJy1GrRDoG/2dsb2JhbABRCYMJMb9GgQEWdIIjAQEBAwEBAQELVwIHCwULC0YnMAYTH4dpBQ26OwSOEIEIMweDAmEDiSKOI5FFgzEc X-IronPort-AV: E=Sophos;i="4.87,945,1363132800"; d="scan'208";a="81844357" Received: from mtv-core-1.cisco.com ([171.68.58.6]) by mtv-iport-3.cisco.com with ESMTP; 26 Jun 2013 15:32:48 +0000 Received: from sjc-vpn5-1626.cisco.com (sjc-vpn5-1626.cisco.com [10.21.94.90]) by mtv-core-1.cisco.com (8.14.5/8.14.5) with ESMTP id r5QFWlJ0031509 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO); Wed, 26 Jun 2013 15:32:47 GMT Content-Type: text/plain; charset=windows-1252 Mime-Version: 1.0 (Mac OS X Mail 6.5 \(1508\)) From: Paul Quinn In-Reply-To: Date: Wed, 26 Jun 2013 08:32:46 -0700 Content-Transfer-Encoding: quoted-printable Message-Id: References: To: Ron Parker X-Mailer: Apple Mail (2.1508) Cc: "nsc@ietf.org" Subject: Re: [nsc] Comments on draft-quinn-nsh-00 X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 15:32:56 -0000 Hi Ron, Thanks for the thoughtful comments! My apologies for not getting back = to you sooner, I've been traveling. See my comments inline. Paul On Jun 24, 2013, at 8:19 AM, Ron Parker = wrote: > I am new to this group and want to thank the authors and participants = for the good work you=92ve done so far. I feel that this work is very = complementary to the Network Function Virtualization Forum=92s nascent = network function chaining work. Draft-quinn-nsh-00 and = draft-quinn-nsc-problem-statement-00 are, however, appropriately more = general in that mobile data services are not assumed to be the de facto = business problem and co-location of network functions are not assumed to = be within the data center. >=20 PQ> Yes. We make no assumption on applicability to particular market = segments or location of service functions =96 all are within scope of = this work.=20 > My specific questions and comments on draft-quinn-nsh-00 follow=85 >=20 > General =96 I agree with some previous comments on the thread that = =93service=94 is an overloaded term. I prefer the NFV=92s terms = =93network function=94 and =93network function chaining=94. Let the = business folks determine what set of functions constitute a =93service=94 = that consumers or businesses will pay for. PQ> We tried use the term service function to describe the specific = "services": e.g. firewall. As Jim Guichard mentioned on the alias: = services are what a customer buys and those services could be = constructed using one or more "service functions". I don't want to use = the term network function as that implies that the network somehow does = something here which is not the case as the function applied is not = necessarily at the network layer.=20 >=20 > 2.2 Problem Statement; 1. Topological Dependencies =96 This would be a = good place to clarify whether service chains are unidirectional or = bidirectional. If bidirectional, is symmetry assumed (i.e. if from = client to server network functions A, B, C are invoked, then from server = to client, is it always C, B, A? Must a classifier be at both ends of = the chain? Must it be the same classifier at each end? >=20 PQ> Point well taken, we have two choices: do you have suggested text = you think captures this, or I'm happy to send out a revision that = incorporates comments to date early next week. One thing to note: = service chains could be unidirectional, bidirectional or something = in-between and that symmetry may or may not be required. As part of the = work we will have to define the "types" of service chains and this is = something that the WG would be assigned to do.=20 > 2.2 Problem Statement =96 It may be useful to discuss how standard IP = routing and NSC co-exist. For example, a deployment based on this = technique may utilize IP routing to ensure packets arrive at the two = ends of a service chain, but the service overlay is used to steer = traffic between the two ends of the chain. >=20 PQ> Can you elaborate a bit more? In general, routing carries the = encapsulated packets from service to service. The chaining function = provides the service function destination for routing lookup. IP = routing is simply used for reachability of each service function within = a service chain.=20 > 3.1 NSH Actions; 1. Insert / Remove Service Header =96 this section = introduces the concept of control plane to indicate which node is =93end = of the chain=94. This could also be accomplished via the management = plane. Perhaps a discussion of control plane and management plane = could be added to section 2.2? =46rom the balance of the draft, it = seems that control plane is deferred for a future proposal? >=20 PQ> Precisely. We expect there to be several possible control plane = options depending on environment and use case and this is something the = WG would be assigned to work through. > 3.1 NSH Actions; 3. Update a Service Header =96 this section = introduces the concept of a service index. =46rom context, I assume = that this indicates the previous node was considered to be index N out = of M total nodes? Is it 0-based or 1-based? PQ> Correct, the index provides location with the path, and is = decremented. Service index=3D0 means the packet should be discarded. >=20 > 3.3 NSH Usage; 2. Service Chaining =96 this section introduces OAM = messages, but none are defined elsewhere in this document. Does this = lend itself to a separate draft? If so, should this document = acknowledge that it won=92t attempt to define OAM procedures? Perhaps = it could acknowledge the possibility that the fields following the Base = Header would be different for OAM messages? >=20 PQ> Yes, the plan is to have a future OAM focused document. I'll = update the draft to reflect that. > 3.3 NSH Usage; 3. Metadata Sharing =96It is mentioned that semantics = of the metadata require coordination via a control plane, but this seems = somewhat contrary to the previous statement that a fixed size and format = header is used for simplicity. PQ> The size is fixed, but the semantics of the data itself needs to be = coordinated by the control plane. For example, if the network shared = context header is used to convey some form of classification, the value = carried in the context but be meaningful to all participating nodes. = The control plane assigns that meaning.=20 >=20 > 3.4 NSH Proxy Nodes =96 it is mentioned that the proxy may deliver the = packet to the service node via a local attachment circuit. I suggest = that it should be specified that the service node is required to deliver = the packet back to the proxy (i.e., =93bump-in-the-wire=94 forwarding = paradigm) via the same or different access circuit. Also, this = mechanism could also be extended to non-NSH tunnels (i..e, IP-in-IP or = GRE). >=20 PQ> Good clarifications, and yes, the same of different circuits is = viable, as are other encaps. I'll update the doc. > 4. Header Format; C bit =96 the text states that when C bit is set, = there are one or more contexts present. But then states that the = format is fixed to 4 32-bit contexts. PQ> This is a parsing optimization: there are always 4 contexts but if = non are used, then c=3D0 and they don't need to be checked. >=20 > 4. Header Format =96 the distinction of network/service and = platform/shared is not clear to me. Perhaps some usage examples could = be added to help clarify? >=20 PQ> Very valid comment, the next revision will include examples. > 4. Header Format =96 for deployment topologies that encompass access = networks (e.g., most), it is very common to wish to convey a subscriber = identity (e.g., IMSI or MSISDN in a mobile network) distinct from the IP = address used in the packet. This may facilitate application of local = policy. It can also facilitate application of local policy in network = functions chained after a NAT function. Perhaps an optional 128-bit = generic subscriber identity can be considered in the NSH? On the = other hand, I can see where keeping the NSH size minimal is extremely = beneficial. One thought here would be a mechanism that could be = employed by stateful network functions using flag bits to acknowledge = that certain metadata has been received and is no longer necessary in = subsequent packets of the fully qualified flow. With this approach, = the 128-bit generic subscriber identity could be conveyed in the first N = packets to the next node until a packet is received back from the next = node with the corresponding acknowledgement bit set. This approach = could be extended to other fields, including the 4 context fields = already defined. >=20 > =20 PQ> The header is flexible enough to accommodate such a case but I need = to understand the use case in more detail. In theory, a control plane = could indicate that the context headers are to be interpreted = differently. Again, I'd like to get a better sense of how you'd use = this. >=20 > _______________________________________________ > nsc mailing list > nsc@ietf.org > https://www.ietf.org/mailman/listinfo/nsc From Ron_Parker@affirmednetworks.com Wed Jun 26 08:50:51 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 3E89121F842E for ; Wed, 26 Jun 2013 08:50:51 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.599 X-Spam-Level: X-Spam-Status: No, score=-2.599 tagged_above=-999 required=5 tests=[AWL=0.000, BAYES_00=-2.599] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id FL3VrdubnYPe for ; Wed, 26 Jun 2013 08:50:46 -0700 (PDT) Received: from hub021-ca-5.exch021.serverdata.net (hub021-ca-5.exch021.serverdata.net [64.78.56.70]) by ietfa.amsl.com (Postfix) with ESMTP id C9AA411E81CB for ; Wed, 26 Jun 2013 08:50:43 -0700 (PDT) Received: from MBX021-W3-CA-2.exch021.domain.local ([10.254.4.78]) by HUB021-CA-5.exch021.domain.local ([10.254.4.89]) with mapi id 14.03.0123.003; Wed, 26 Jun 2013 08:50:42 -0700 From: Ron Parker To: Paul Quinn Thread-Topic: [nsc] Comments on draft-quinn-nsh-00 Thread-Index: Ac5w7gaFkR09Y/lqTveMLRqlAXe6pQBzw4MAAA54RDA= Date: Wed, 26 Jun 2013 15:50:43 +0000 Message-ID: References: In-Reply-To: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [50.203.66.100] Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Cc: "nsc@ietf.org" Subject: Re: [nsc] Comments on draft-quinn-nsh-00 X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 26 Jun 2013 15:50:51 -0000 Hi, Paul. Thanks for the response. My follow up embedded. BR, Ron -----Original Message----- From: Paul Quinn [mailto:paulq@cisco.com]=20 Sent: Wednesday, June 26, 2013 11:33 AM To: Ron Parker Cc: nsc@ietf.org Subject: Re: [nsc] Comments on draft-quinn-nsh-00 Hi Ron, Thanks for the thoughtful comments! My apologies for not getting back to y= ou sooner, I've been traveling. See my comments inline. Paul On Jun 24, 2013, at 8:19 AM, Ron Parker w= rote: > I am new to this group and want to thank the authors and participants for= the good work you've done so far. I feel that this work is very compleme= ntary to the Network Function Virtualization Forum's nascent network functi= on chaining work. Draft-quinn-nsh-00 and draft-quinn-nsc-problem-statemen= t-00 are, however, appropriately more general in that mobile data services = are not assumed to be the de facto business problem and co-location of netw= ork functions are not assumed to be within the data center. >=20 PQ> Yes. We make no assumption on applicability to particular market segme= nts or location of service functions - all are within scope of this work.=20 > My specific questions and comments on draft-quinn-nsh-00 follow... >=20 > General - I agree with some previous comments on the thread that "service= " is an overloaded term. I prefer the NFV's terms "network function" and = "network function chaining". Let the business folks determine what set of= functions constitute a "service" that consumers or businesses will pay for= . PQ> We tried use the term service function to describe the specific "servi= ces": e.g. firewall. As Jim Guichard mentioned on the alias: services are = what a customer buys and those services could be constructed using one or m= ore "service functions". I don't want to use the term network function as t= hat implies that the network somehow does something here which is not the c= ase as the function applied is not necessarily at the network layer.=20 Ron> Makes sense. >=20 > 2.2 Problem Statement; 1. Topological Dependencies - This would be a good= place to clarify whether service chains are unidirectional or bidirectiona= l. If bidirectional, is symmetry assumed (i.e. if from client to server n= etwork functions A, B, C are invoked, then from server to client, is it alw= ays C, B, A? Must a classifier be at both ends of the chain? Must it be = the same classifier at each end? >=20 PQ> Point well taken, we have two choices: do you have suggested text you = think captures this, or I'm happy to send out a revision that incorporates = comments to date early next week. One thing to note: service chains could = be unidirectional, bidirectional or something in-between and that symmetry = may or may not be required. As part of the work we will have to define the = "types" of service chains and this is something that the WG would be assign= ed to do.=20 Ron> I don't have specific text at this point. Agree that unidirectional,= inversely-symmetrical bidirectional, and asymmetrical bidirectional are al= l valid. Suggest that the flexible location of the classifier(s) be broug= ht out, too. > 2.2 Problem Statement - It may be useful to discuss how standard IP routi= ng and NSC co-exist. For example, a deployment based on this technique ma= y utilize IP routing to ensure packets arrive at the two ends of a service = chain, but the service overlay is used to steer traffic between the two end= s of the chain. >=20 PQ> Can you elaborate a bit more? In general, routing carries the encapsu= lated packets from service to service. The chaining function provides the = service function destination for routing lookup. IP routing is simply use= d for reachability of each service function within a service chain.=20 Ron> I feel that many people thinking about this topic have somehow arrived= at a conclusion that OpenFlow is automatically and exclusively the basis f= or achieving things like NSC. > 3.1 NSH Actions; 1. Insert / Remove Service Header - this section introdu= ces the concept of control plane to indicate which node is "end of the chai= n". This could also be accomplished via the management plane. Perhaps a= discussion of control plane and management plane could be added to section= 2.2? From the balance of the draft, it seems that control plane is defer= red for a future proposal? >=20 PQ> Precisely. We expect there to be several possible control plane optio= ns depending on environment and use case and this is something the WG would= be assigned to work through. Ron> My suggestion is that the concept be explicitly supportable with a co= ntrol plane and without (i.e., solely via coordinated management of the aff= ected Network Elements). This may allow for earlier adoptions and thereby= market validation of the concepts. > 3.1 NSH Actions; 3. Update a Service Header - this section introduces the= concept of a service index. From context, I assume that this indicates t= he previous node was considered to be index N out of M total nodes? Is it= 0-based or 1-based? PQ> Correct, the index provides location with the path, and is decremented= . Service index=3D0 means the packet should be discarded. Ron> I suggest that the text be enhanced to bring out the semantic meaning = more explicitly. >=20 > 3.3 NSH Usage; 2. Service Chaining - this section introduces OAM messages= , but none are defined elsewhere in this document. Does this lend itself = to a separate draft? If so, should this document acknowledge that it won'= t attempt to define OAM procedures? Perhaps it could acknowledge the pos= sibility that the fields following the Base Header would be different for O= AM messages? >=20 PQ> Yes, the plan is to have a future OAM focused document. I'll update t= he draft to reflect that. > 3.3 NSH Usage; 3. Metadata Sharing -It is mentioned that semantics of the= metadata require coordination via a control plane, but this seems somewhat= contrary to the previous statement that a fixed size and format header is = used for simplicity. PQ> The size is fixed, but the semantics of the data itself needs to be coo= rdinated by the control plane. For example, if the network shared context = header is used to convey some form of classification, the value carried in = the context but be meaningful to all participating nodes. The control plan= e assigns that meaning.=20 Ron> Suggest that the concept include the option whereby the management pla= ne assigns that meaning, too. >=20 > 3.4 NSH Proxy Nodes - it is mentioned that the proxy may deliver the pack= et to the service node via a local attachment circuit. I suggest that it s= hould be specified that the service node is required to deliver the packet = back to the proxy (i.e., "bump-in-the-wire" forwarding paradigm) via the sa= me or different access circuit. Also, this mechanism could also be extend= ed to non-NSH tunnels (i..e, IP-in-IP or GRE). >=20 PQ> Good clarifications, and yes, the same of different circuits is viable= , as are other encaps. I'll update the doc. > 4. Header Format; C bit - the text states that when C bit is set, there a= re one or more contexts present. But then states that the format is fixed= to 4 32-bit contexts. PQ> This is a parsing optimization: there are always 4 contexts but if non= are used, then c=3D0 and they don't need to be checked. Ron> This was more of an editorial comment. "When the C bit is set, all f= our contexts are present" would more accurately reflect the intention, if I= have interpreted correctly. >=20 > 4. Header Format - the distinction of network/service and platform/shared= is not clear to me. Perhaps some usage examples could be added to help c= larify? >=20 PQ> Very valid comment, the next revision will include examples. > 4. Header Format - for deployment topologies that encompass access netw= orks (e.g., most), it is very common to wish to convey a subscriber identit= y (e.g., IMSI or MSISDN in a mobile network) distinct from the IP address u= sed in the packet. This may facilitate application of local policy. It = can also facilitate application of local policy in network functions chaine= d after a NAT function. Perhaps an optional 128-bit generic subscriber id= entity can be considered in the NSH? On the other hand, I can see where = keeping the NSH size minimal is extremely beneficial. One thought here wo= uld be a mechanism that could be employed by stateful network functions usi= ng flag bits to acknowledge that certain metadata has been received and is = no longer necessary in subsequent packets of the fully qualified flow. Wi= th this approach, the 128-bit generic subscriber identity could be conveyed= in the first N packets to the next node until a packet is received back fr= om the next node with the corresponding acknowledgement bit set. This app= roach could be extended to other fields, including the 4 context fields alr= eady defined. >=20 > =20 PQ> The header is flexible enough to accommodate such a case but I need to = understand the use case in more detail. In theory, a control plane could i= ndicate that the context headers are to be interpreted differently. Again,= I'd like to get a better sense of how you'd use this. Ron> Given this split notion of classification-capable nodes and non-classi= fication-capable nodes, could the classification nodes add more value than = simply choosing which SF's need to be invoked. In particular, certain SF'= s would be capable of providing differentiated services for different group= s of subscribers. Can the classifier determine not only that some SF need= s to be inserted but also which locally significant set of policy should be= applied? If so, how is this conveyed from the classifier to the SF? In= band? Out of band? >=20 > _______________________________________________ > nsc mailing list > nsc@ietf.org > https://www.ietf.org/mailman/listinfo/nsc From christian.jacquenet@orange.com Thu Jun 27 04:41:01 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id CEAC621F9CD4 for ; Thu, 27 Jun 2013 04:41:01 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.247 X-Spam-Level: X-Spam-Status: No, score=-2.247 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HELO_EQ_FR=0.35, HTML_MESSAGE=0.001, UNPARSEABLE_RELAY=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Fkptd-egQ4H6 for ; Thu, 27 Jun 2013 04:40:56 -0700 (PDT) Received: from relais-inet.francetelecom.com (relais-ias91.francetelecom.com [193.251.215.91]) by ietfa.amsl.com (Postfix) with ESMTP id A385621F9CDF for ; Thu, 27 Jun 2013 04:40:49 -0700 (PDT) Received: from omfedm08.si.francetelecom.fr (unknown [xx.xx.xx.4]) by omfedm14.si.francetelecom.fr (ESMTP service) with ESMTP id BBD8E22D5C1; Thu, 27 Jun 2013 13:40:48 +0200 (CEST) Received: from PUEXCH41.nanterre.francetelecom.fr (unknown [10.101.44.30]) by omfedm08.si.francetelecom.fr (ESMTP service) with ESMTP id 9C100238056; Thu, 27 Jun 2013 13:40:48 +0200 (CEST) Received: from PUEXCB1C.nanterre.francetelecom.fr ([10.101.44.8]) by PUEXCH41.nanterre.francetelecom.fr ([10.101.44.30]) with mapi; Thu, 27 Jun 2013 13:40:48 +0200 From: To: Ron Parker , "nsc@ietf.org" Date: Thu, 27 Jun 2013 13:40:47 +0200 Thread-Topic: Comments on draft-boucadair-network-function-chaining-00 Thread-Index: Ac5x6/wxcUX+ReJaT8eNlB4kpr2L6QASIKPAABD+vdAALFBOMA== Message-ID: <21207_1372333248_51CC24C0_21207_4730_1_983A1D8DA0DA5F4EB747BF34CBEE5CD15C435E8C74@PUEXCB1C.nanterre.francetelecom.fr> References: <32603_1372228737_51CA8C81_32603_2651_1_983A1D8DA0DA5F4EB747BF34CBEE5CD15C418F0433@PUEXCB1C.nanterre.francetelecom.fr> In-Reply-To: Accept-Language: fr-FR Content-Language: fr-FR X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: fr-FR Content-Type: multipart/alternative; boundary="_000_983A1D8DA0DA5F4EB747BF34CBEE5CD15C435E8C74PUEXCB1Cnante_" MIME-Version: 1.0 X-PMX-Version: 5.6.1.2065439, Antispam-Engine: 2.7.2.376379, Antispam-Data: 2013.6.27.111232 Subject: Re: [nsc] Comments on draft-boucadair-network-function-chaining-00 X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 27 Jun 2013 11:41:01 -0000 --_000_983A1D8DA0DA5F4EB747BF34CBEE5CD15C435E8C74PUEXCB1Cnante_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Hello Ron, A few additional comments. [snip] [CJ] I think it's primarily a design choice that will be conditioned by the= very nature of the NLF to be activated. I'm not sure each and every NLF we= can think about has implicit load balancing capabilities. Therefore, I wou= ld consider load balancing as a possible (but not necessarily mandatory) ob= jective of the NLFC machinery. [Ron] If the NLFC explicitly allows for load balancing by the PDP, then som= e requirements would follow wrt availability determination of the constitue= nt addressable entities. Optionally, a means to determine loading of the c= onstituent addressable entities so that the PDP may make a more informed de= cision. For the first, ping and BFD could be considered. For the second= , SNMP could be considered. [CJ] I'm not sure a Ping or BFD would be sufficient to help the PDP make a = decision about load balancing. I think that what you consider as an option = is actually mandatory for the sake of efficient load balancing. SNMP is one= candidate protocol to report about the loading conditions of an addressabl= e entity indeed. There are others, e.g., COPS-PR along with the support of = RFC 3571. Is it an objective for NLFC to determine the availability of an NLF? [CJ] No, I think the availability of an NLF suggests another mechanism, e.g= ., the (automated) discovery of the network topology, devices and their res= pective capabilities which will feed the PDP intelligence with information = to not only dynamically structure a network service but also to trigger NLF= C according to the outcomes of the negotiation of the service parameters be= tween the customer and the provider, for example. [Ron] End-to-end reliability and resiliency could be enhanced if the PDP co= uld prevent launching packets into a broken chain. Perhaps in some cases,= an unreachable NLF should be pruned from the chain while in other cases, t= he traffic should be dropped. A trap or other indication coming from the = PDP would assist in overall network visibility and troubleshooting. [CJ] I guess you meant a trap coming from the NLF-enabled device, right? I = agree that the NLFC behavior is very likely to depend on the nature of the = NLF functions to be invoked as per the NLFC Map, possibly yielding packet d= ropping (but this is a situation that should be avoided: this brings back t= he kind of feedback mechanism you rightfully mention in the previous commen= t so that the PDP can make a decision accordingly, which could include the = triggering of either a new resource provisioning cycle, a new negotiation p= hase with the customer, the computation of another differentiated forwardin= g path, or a combination thereof. Cheers, Christian. ___________________________________________________________________________= ______________________________________________ Ce message et ses pieces jointes peuvent contenir des informations confiden= tielles ou privilegiees et ne doivent donc pas etre diffuses, exploites ou copies sans autorisation. Si vous avez recu= ce message par erreur, veuillez le signaler a l'expediteur et le detruire ainsi que les pieces jointes. Les messages el= ectroniques etant susceptibles d'alteration, Orange decline toute responsabilite si ce message a ete altere, deforme ou = falsifie. Merci. This message and its attachments may contain confidential or privileged inf= ormation that may be protected by law; they should not be distributed, used or copied without authorisation. If you have received this email in error, please notify the sender and dele= te this message and its attachments. As emails may be altered, Orange is not liable for messages that have been = modified, changed or falsified. Thank you. --_000_983A1D8DA0DA5F4EB747BF34CBEE5CD15C435E8C74PUEXCB1Cnante_ Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable

Hello Ron,=

 

A few additional comments.

[snip]

[CJ] I think it’s primarily a desi= gn choice that will be conditioned by the very nature of the NLF to be acti= vated. I’m not sure each and every NLF we can think about has implici= t load balancing capabilities. Therefore, I would consider load balancing a= s a possible (but not necessarily mandatory) objective of the NLFC machiner= y.

<= span style=3D'color:#1F497D'>[Ron] If the NLFC explicitly allows for load b= alancing by the PDP, then some requirements would follow wrt availability d= etermination of the constituent addressable entities.  Optionally, a m= eans to determine loading of the constituent addressable entities so that t= he PDP may make a more informed decision.   For the first, ping a= nd BFD could be considered.   For the second, SNMP could be consi= dered.

[CJ] I’m not sure a Pi= ng or BFD would be sufficient to help the PDP make a decision about load ba= lancing. I think that what you consider as an option is actually mandatory = for the sake of efficient load balancing. SNMP is one candidate protocol to= report about the loading conditions of an addressable entity indeed. There= are others, e.g., COPS-PR along with the support of RFC 3571.

 

Is it an objective for NLFC to determine the availability = of an NLF?

[CJ] No, I think the availability of an = NLF suggests another mechanism, e.g., the (automated) discovery of the netw= ork topology, devices and their respective capabilities which will feed the= PDP intelligence with information to not only dynamically structure a netw= ork service but also to trigger NLFC according to the outcomes of the negot= iation of the service parameters between the customer and the provider, for= example.

[Ron] End-to-end reliability and resili= ency could be enhanced if the PDP could prevent launching packets into a br= oken chain.   Perhaps in some cases, an unreachable NLF should be= pruned from the chain while in other cases, the traffic should be dropped.=    A trap or other indication coming from the PDP would assist in= overall network visibility and troubleshooting.

=
 
<= pre>[CJ] I guess you meant a trap coming from= the NLF-enabled device, right? I agree that the NLFC behavior is very like= ly to depend on the nature of the NLF functions to be invoked as per the NL= FC Map, possibly yielding packet dropping (but this is a situation that sho= uld be avoided: this brings back the kind of feedback mechanism you rightfu= lly mention in the previous comment so that the PDP can make a decision acc= ordingly, which could include the triggering of either a new resource provi= sioning cycle, a new negotiation phase with the customer, the computation o= f another differentiated forwarding path, or a combination thereof.
 =

Cheers,
 
Christian.
________=
___________________________________________________________________________=
______________________________________

Ce message et ses pieces jointes peuvent contenir des informations confiden=
tielles ou privilegiees et ne doivent donc
pas etre diffuses, exploites ou copies sans autorisation. Si vous avez recu=
 ce message par erreur, veuillez le signaler
a l'expediteur et le detruire ainsi que les pieces jointes. Les messages el=
ectroniques etant susceptibles d'alteration,
Orange decline toute responsabilite si ce message a ete altere, deforme ou =
falsifie. Merci.

This message and its attachments may contain confidential or privileged inf=
ormation that may be protected by law;
they should not be distributed, used or copied without authorisation.
If you have received this email in error, please notify the sender and dele=
te this message and its attachments.
As emails may be altered, Orange is not liable for messages that have been =
modified, changed or falsified.
Thank you.
= --_000_983A1D8DA0DA5F4EB747BF34CBEE5CD15C435E8C74PUEXCB1Cnante_-- From narten@us.ibm.com Thu Jun 27 06:30:11 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 055A321F9DDB for ; Thu, 27 Jun 2013 06:30:11 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -110.599 X-Spam-Level: X-Spam-Status: No, score=-110.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8, USER_IN_WHITELIST=-100] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id NqCc8WflYi8Z for ; Thu, 27 Jun 2013 06:30:05 -0700 (PDT) Received: from e39.co.us.ibm.com (e39.co.us.ibm.com [32.97.110.160]) by ietfa.amsl.com (Postfix) with ESMTP id D16FF21F9DD0 for ; Thu, 27 Jun 2013 06:30:04 -0700 (PDT) Received: from /spool/local by e39.co.us.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Thu, 27 Jun 2013 07:30:03 -0600 Received: from d01dlp03.pok.ibm.com (9.56.250.168) by e39.co.us.ibm.com (192.168.1.139) with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted; Thu, 27 Jun 2013 07:30:00 -0600 Received: from d01relay02.pok.ibm.com (d01relay02.pok.ibm.com [9.56.227.234]) by d01dlp03.pok.ibm.com (Postfix) with ESMTP id 58F4EC9003E for ; Thu, 27 Jun 2013 09:29:58 -0400 (EDT) Received: from d01av03.pok.ibm.com (d01av03.pok.ibm.com [9.56.224.217]) by d01relay02.pok.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id r5RDTxvN260712 for ; Thu, 27 Jun 2013 09:29:59 -0400 Received: from d01av03.pok.ibm.com (loopback [127.0.0.1]) by d01av03.pok.ibm.com (8.14.4/8.13.1/NCO v10.0 AVout) with ESMTP id r5RDTwXK024713 for ; Thu, 27 Jun 2013 10:29:58 -0300 Received: from cichlid.raleigh.ibm.com ([9.49.218.175]) by d01av03.pok.ibm.com (8.14.4/8.13.1/NCO v10.0 AVin) with ESMTP id r5RDTvD8024628 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 27 Jun 2013 10:29:58 -0300 Received: from cichlid.raleigh.ibm.com (localhost [127.0.0.1]) by cichlid.raleigh.ibm.com (8.14.4/8.12.5) with ESMTP id r5RDTuak003634; Thu, 27 Jun 2013 09:29:57 -0400 Message-Id: <201306271329.r5RDTuak003634@cichlid.raleigh.ibm.com> To: Robert Raszuk In-reply-to: References: Comments: In-reply-to Robert Raszuk message dated "Sun, 23 Jun 2013 22:07:55 +0200." Date: Thu, 27 Jun 2013 09:29:56 -0400 From: Thomas Narten X-TM-AS-MML: No X-Content-Scanned: Fidelis XPS MAILER x-cbid: 13062713-3620-0000-0000-0000034F72F0 Cc: "nsc@ietf.org" Subject: Re: [nsc] Network Service Header - draft-quinn-nsh-00.txt X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 27 Jun 2013 13:30:11 -0000 Robert Raszuk writes: > I have a basic question reg Network Service Header draft. > Don't we already have sufficient room in v6 header (including it's > extension header) to embed services in it ? Do we need to invent new > header each time someone proposes a new approach ;) ? IMO, discussions about how best to encode stuff is a tad premature. We should focus discussion on whether some item of information needs to be included in a packet or not, and the motivation/pros/cons of doing that. > Just like we can get service discovery accomplished for free when we > embed service in its reachability advertisement: > http://telematics.tm.kit.edu/publications/Files/491/ipv6-app-contest-2011-paper.pdf "service discovery for free" is an oxymoron. There are a ton of reasons why the above proposal is not a good idea. Encoding additional semantics into IP addresses is fraught with difficulties, and except in very limited situations, the IETF has not been willing to do so. Thomas From Ron_Parker@affirmednetworks.com Thu Jun 27 07:18:26 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 2CDED21F9E32 for ; Thu, 27 Jun 2013 07:18:26 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.598 X-Spam-Level: X-Spam-Status: No, score=-2.598 tagged_above=-999 required=5 tests=[AWL=-0.000, BAYES_00=-2.599, HTML_MESSAGE=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id TVu4Gvzm70mY for ; Thu, 27 Jun 2013 07:18:20 -0700 (PDT) Received: from hub021-ca-4.exch021.serverdata.net (hub021-ca-4.exch021.serverdata.net [64.78.22.171]) by ietfa.amsl.com (Postfix) with ESMTP id 35EC321F9E30 for ; Thu, 27 Jun 2013 07:18:20 -0700 (PDT) Received: from MBX021-W3-CA-2.exch021.domain.local ([10.254.4.78]) by HUB021-CA-4.exch021.domain.local ([10.254.4.39]) with mapi id 14.03.0123.003; Thu, 27 Jun 2013 07:18:18 -0700 From: Ron Parker To: "christian.jacquenet@orange.com" , "nsc@ietf.org" Thread-Topic: Comments on draft-boucadair-network-function-chaining-00 Thread-Index: Ac5x6/wxcUX+ReJaT8eNlB4kpr2L6QASIKPAABD+vdAALFBOMAAFlIig Date: Thu, 27 Jun 2013 14:18:18 +0000 Message-ID: References: <32603_1372228737_51CA8C81_32603_2651_1_983A1D8DA0DA5F4EB747BF34CBEE5CD15C418F0433@PUEXCB1C.nanterre.francetelecom.fr> <21207_1372333248_51CC24C0_21207_4730_1_983A1D8DA0DA5F4EB747BF34CBEE5CD15C435E8C74@PUEXCB1C.nanterre.francetelecom.fr> In-Reply-To: <21207_1372333248_51CC24C0_21207_4730_1_983A1D8DA0DA5F4EB747BF34CBEE5CD15C435E8C74@PUEXCB1C.nanterre.francetelecom.fr> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [50.203.66.100] Content-Type: multipart/alternative; boundary="_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66A519MBX021W3CA2exch_" MIME-Version: 1.0 Subject: Re: [nsc] Comments on draft-boucadair-network-function-chaining-00 X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 27 Jun 2013 14:18:26 -0000 --_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66A519MBX021W3CA2exch_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Thanks for the continued discussion, Christian. Regarding load balancing performed by PDP, there may be simple scenarios wh= ere live connectivity to an NLF instance within a pool of NLF instances is = a sufficient condition in order to send additional work to it and I don't w= ant to preclude that scenario. More complex scenarios would, of course, r= equire some way for the PDP to obtain a more sophisticated view of loading= state of each of the NLF instances within the pool of NLF instances. Determination of liveness is important, IMO, even for invocation of non-loa= d balanced NLFs by the PDP. Given that the PDP is an "orchestrator" of en= d-to-end service chains, it is a natural entity to query as to a high level= view of the health of those chains. This would also facilitate support o= f active/standby NLF instances, again, outside of load balancing. My suggestion is that liveness determination between PDP and NLF be explici= tly part of the architecture. Thanks. Ron From: christian.jacquenet@orange.com [mailto:christian.jacquenet@orange.com= ] Sent: Thursday, June 27, 2013 7:41 AM To: Ron Parker; nsc@ietf.org Subject: RE: Comments on draft-boucadair-network-function-chaining-00 Hello Ron, A few additional comments. [snip] [CJ] I think it's primarily a design choice that will be conditioned by the= very nature of the NLF to be activated. I'm not sure each and every NLF we= can think about has implicit load balancing capabilities. Therefore, I wou= ld consider load balancing as a possible (but not necessarily mandatory) ob= jective of the NLFC machinery. [Ron] If the NLFC explicitly allows for load balancing by the PDP, then som= e requirements would follow wrt availability determination of the constitue= nt addressable entities. Optionally, a means to determine loading of the c= onstituent addressable entities so that the PDP may make a more informed de= cision. For the first, ping and BFD could be considered. For the second= , SNMP could be considered. [CJ] I'm not sure a Ping or BFD would be sufficient to help the PDP make a = decision about load balancing. I think that what you consider as an option = is actually mandatory for the sake of efficient load balancing. SNMP is one= candidate protocol to report about the loading conditions of an addressabl= e entity indeed. There are others, e.g., COPS-PR along with the support of = RFC 3571. Is it an objective for NLFC to determine the availability of an NLF? [CJ] No, I think the availability of an NLF suggests another mechanism, e.g= ., the (automated) discovery of the network topology, devices and their res= pective capabilities which will feed the PDP intelligence with information = to not only dynamically structure a network service but also to trigger NLF= C according to the outcomes of the negotiation of the service parameters be= tween the customer and the provider, for example. [Ron] End-to-end reliability and resiliency could be enhanced if the PDP co= uld prevent launching packets into a broken chain. Perhaps in some cases,= an unreachable NLF should be pruned from the chain while in other cases, t= he traffic should be dropped. A trap or other indication coming from the = PDP would assist in overall network visibility and troubleshooting. [CJ] I guess you meant a trap coming from the NLF-enabled device, right? I = agree that the NLFC behavior is very likely to depend on the nature of the = NLF functions to be invoked as per the NLFC Map, possibly yielding packet d= ropping (but this is a situation that should be avoided: this brings back t= he kind of feedback mechanism you rightfully mention in the previous commen= t so that the PDP can make a decision accordingly, which could include the = triggering of either a new resource provisioning cycle, a new negotiation p= hase with the customer, the computation of another differentiated forwardin= g path, or a combination thereof. Cheers, Christian. ___________________________________________________________________________= ______________________________________________ Ce message et ses pieces jointes peuvent contenir des informations confiden= tielles ou privilegiees et ne doivent donc pas etre diffuses, exploites ou copies sans autorisation. Si vous avez recu= ce message par erreur, veuillez le signaler a l'expediteur et le detruire ainsi que les pieces jointes. Les messages el= ectroniques etant susceptibles d'alteration, Orange decline toute responsabilite si ce message a ete altere, deforme ou = falsifie. Merci. This message and its attachments may contain confidential or privileged inf= ormation that may be protected by law; they should not be distributed, used or copied without authorisation. If you have received this email in error, please notify the sender and dele= te this message and its attachments. As emails may be altered, Orange is not liable for messages that have been = modified, changed or falsified. Thank you. --_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66A519MBX021W3CA2exch_ Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable

Thanks for the continu= ed discussion, Christian.

 

Regarding load balanci= ng performed by PDP, there may be simple scenarios where live connectivity = to an NLF instance within a pool of NLF instances is a sufficient condition= in order to send additional work to it and I don’t want to preclude that scenario.   More comp= lex scenarios would, of course, require some way for the PDP to obtain &nbs= p;a more sophisticated view of loading state of each of the NLF instances w= ithin the pool of NLF instances.

 

Determination of liven= ess is important, IMO, even for invocation of non-load balanced NLFs by the= PDP.   Given that the PDP is an “orchestrator” of en= d-to-end service chains, it is a natural entity to query as to a high level view of the health of those chains.   This would= also facilitate support of active/standby NLF instances, again, outside of= load balancing.

 

My suggestion is that = liveness determination between PDP and NLF be explicitly part of the archit= ecture.

 

Thanks.

 

   Ron<= /o:p>

 

 

From: christian.jacquenet@orange.com [mailto:= christian.jacquenet@orange.com]
Sent: Thursday, June 27, 2013 7:41 AM
To: Ron Parker; nsc@ietf.org
Subject: RE: Comments on draft-boucadair-network-function-chaining-0= 0

 

Hello Ron,

 

A few additional comments.=

[snip]

[CJ] I think it&#= 8217;s primarily a design choice that will be conditioned by the very natur= e of the NLF to be activated. I’m not sure each and every NLF we can think about has implicit load balancing capabilities. Therefore= , I would consider load balancing as a possible (but not necessarily mandat= ory) objective of the NLFC machinery.

[Ron] If the NLFC explicitly allows for load balancing by the PDP, the= n some requirements would follow wrt availability determination of the cons= tituent addressable entities.  Optionally, a means to determine loading of the constituent addressable entities so th= at the PDP may make a more informed decision.   For the first, pi= ng and BFD could be considered.   For the second, SNMP could be c= onsidered.

[CJ] I’m not sure a Ping or BFD would = be sufficient to help the PDP make a decision about load balancing. I think= that what you consider as an option is actually mandatory for the sake of efficient load balancing. SNMP is one candidate protocol t= o report about the loading conditions of an addressable entity indeed. Ther= e are others, e.g., COPS-PR along with the support of RFC 3571.<= /span>

 

Is it an objective for N= LFC to determine the availability of an NLF?<= o:p>

[CJ] No, I think = the availability of an NLF suggests another mechanism, e.g., the (automated= ) discovery of the network topology, devices and their respective capabilities which will feed the PDP intelligence with in= formation to not only dynamically structure a network service but also to t= rigger NLFC according to the outcomes of the negotiation of the service par= ameters between the customer and the provider, for example.

[Ron] End-to-end reliability and resiliency could be enhanced if the P= DP could prevent launching packets into a broken chain.   Perhaps= in some cases, an unreachable NLF should be pruned from the chain while in other cases, the traffic should be dropped. &= nbsp; A trap or other indication coming from the PDP would assist in overal= l network visibility and troubleshooting.

 
[CJ] I guess you meant a trap coming from the NLF-enabled devi=
ce, right? I agree that the NLFC behavior is very likely to depend on the n=
ature of the NLF functions to be invoked as per the NLFC Map, possibly yiel=
ding packet dropping (but this is a situation that should be avoided: this =
brings back the kind of feedback mechanism you rightfully mention in the pr=
evious comment so that the PDP can make a decision accordingly, which could=
 include the triggering of either a new resource provisioning cycle, a new =
negotiation phase with the customer, the computation of another differentia=
ted forwarding path, or a combination thereof.
 
Cheers,
 
Christian.
_=
___________________________________________________________________________=
_____________________________________________
<=
o:p> 
C=
e message et ses pieces jointes peuvent contenir des informations confident=
ielles ou privilegiees et ne doivent donc
p=
as etre diffuses, exploites ou copies sans autorisation. Si vous avez recu =
ce message par erreur, veuillez le signaler
a=
 l'expediteur et le detruire ainsi que les pieces jointes. Les messages ele=
ctroniques etant susceptibles d'alteration,
O=
range decline toute responsabilite si ce message a ete altere, deforme ou f=
alsifie. Merci.
<=
o:p> 
T=
his message and its attachments may contain confidential or privileged info=
rmation that may be protected by law;
t=
hey should not be distributed, used or copied without authorisation.
I=
f you have received this email in error, please notify the sender and delet=
e this message and its attachments.
A=
s emails may be altered, Orange is not liable for messages that have been m=
odified, changed or falsified.
T=
hank you.
--_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66A519MBX021W3CA2exch_-- From Ron_Parker@affirmednetworks.com Thu Jun 27 11:58:56 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 32FD321F9A3F for ; Thu, 27 Jun 2013 11:58:56 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: 0.003 X-Spam-Level: X-Spam-Status: No, score=0.003 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HTML_MESSAGE=0.001, HTML_OBFUSCATE_10_20=2.601] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id aJunSeGsrR20 for ; Thu, 27 Jun 2013 11:58:50 -0700 (PDT) Received: from hub021-ca-2.exch021.serverdata.net (hub021-ca-2.exch021.serverdata.net [64.78.22.169]) by ietfa.amsl.com (Postfix) with ESMTP id 4CE9621F9EA5 for ; Thu, 27 Jun 2013 11:58:47 -0700 (PDT) Received: from MBX021-W3-CA-2.exch021.domain.local ([10.254.4.78]) by HUB021-CA-2.exch021.domain.local ([10.254.4.33]) with mapi id 14.03.0123.003; Thu, 27 Jun 2013 11:58:45 -0700 From: Ron Parker To: "nsc@ietf.org" Thread-Topic: Question regarding topology for bidirectional service chains Thread-Index: Ac5zaFiFaRP0gdxtSJ2ZgwSmZwBg6Q== Date: Thu, 27 Jun 2013 18:58:45 +0000 Message-ID: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [50.203.66.100] Content-Type: multipart/alternative; boundary="_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66C538MBX021W3CA2exch_" MIME-Version: 1.0 Subject: [nsc] Question regarding topology for bidirectional service chains X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 27 Jun 2013 18:58:56 -0000 --_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66C538MBX021W3CA2exch_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable I have a question for the group regarding bidirectional service chains. For sake of discussion, I will call the directions of packet flow East and = West. Let's also assume that classification-only nodes exist at each end= of the chain and that service-only nodes exist in the middle. And for si= mplicity of discussion, let's say that service invocation order is an exact= mirror in each direction. <---->classifier-West <---->service-node-1 <---->service-node-2 <---->= service-node-3 <---->classifier-East <----> In the eastbound direction, classifier-West first receives the traffic from= some upstream node and determines the service-chain. Likewise, in the we= stbound direction, classifier-East first receives the traffic from some ups= tream node and determines the service-chain. But, what about at the final= end of each chain? Must eastbound traffic traverse classifier-East and = must westbound traffic traverse classifier-West? I'm suspecting that the answer is no, but it does create the asymmetrical r= equirement that a non-classifier node can never be NLF-ingress, but can be = NLF-egress. -->classifier-West -->service-node-1 -->service-node-2 -->service-node= -3 --> <--service-node-1 <--service-node-2 <--service-node-3 <--classifier-Ea= st <-- Thanks. Ron --_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66C538MBX021W3CA2exch_ Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable

I have a question for the group regarding bidirectio= nal service chains.  

 

For sake of discussion, I will call the directions o= f packet flow East and West.    Let’s also assume that= classification-only nodes exist at each end of the chain and that service-= only nodes exist in the middle.   And for simplicity of discussion, let’s say that service invocation order is an exact mirr= or in each direction.

 

     ßàclassifier-West ßàservice-node-1 ßàservice-node-2 ßàservice-node-3 ßàclassifier-East ßà

 

In the eastbound direction, classifier-West first re= ceives the traffic from some upstream node and determines the service-chain= .   Likewise, in the westbound direction, classifier-East first r= eceives the traffic from some upstream node and determines the service-chain.   But, what about at the final= end of each chain?    Must eastbound traffic traverse class= ifier-East and must westbound traffic traverse classifier-West?<= /p>

 

I’m suspecting that the answer is no, but it d= oes create the asymmetrical requirement that a non-classifier node can neve= r be NLF-ingress, but can be NLF-egress.

 

     àclassifier-West àservice-node-1 àservice-node-2 à= ;service-node-3 à

     ßservice-node-1 ßservice-node-2 ßservice-node-3 ß<= /span>classifier-East ß

 

 

Thanks.

 

   Ron

--_000_CDF2F015F4429F458815ED2A6C2B6B0B1A66C538MBX021W3CA2exch_-- From jguichar@cisco.com Thu Jun 27 15:49:07 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 30FAB11E810B for ; Thu, 27 Jun 2013 15:49:07 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -7.997 X-Spam-Level: X-Spam-Status: No, score=-7.997 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HTML_MESSAGE=0.001, HTML_OBFUSCATE_10_20=2.601, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ByxBve-Hw+bC for ; Thu, 27 Jun 2013 15:49:01 -0700 (PDT) Received: from rcdn-iport-4.cisco.com (rcdn-iport-4.cisco.com [173.37.86.75]) by ietfa.amsl.com (Postfix) with ESMTP id 88BAB11E8114 for ; Thu, 27 Jun 2013 15:48:57 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=8957; q=dns/txt; s=iport; t=1372373337; x=1373582937; h=from:to:subject:date:message-id:in-reply-to:mime-version; bh=DRgY+cSQcINzJ3j9QvmXFaE5MbvSLK3+9lVChCuZQME=; b=IKr8HGcQgVALIyzUNSQYBfWsFb/LgHNIW9cyPeu96mjN9nWt2Wo4rfF1 RhgCKGwolgbxK4d9/75BfJieqFYsfBBfajWbrdEnO7SYZOlOmntAXL9Rj U0zZLQDHNsENIYt84MxIZlDZyOwaMiIALYRkZbcPnYzyglW1j/2/86nVH c=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: Ah0FAIfAzFGtJXHA/2dsb2JhbABbgkVEMUm/DIECFnSCIwEBAQEDJwZeAQgRAwECCx05FAkIAQEEARIIiAa7So8kDRMYgwJjA5NzjgGHFoMRgig X-IronPort-AV: E=Sophos;i="4.87,955,1363132800"; d="scan'208,217";a="228456571" Received: from rcdn-core2-5.cisco.com ([173.37.113.192]) by rcdn-iport-4.cisco.com with ESMTP; 27 Jun 2013 22:48:56 +0000 Received: from xhc-aln-x06.cisco.com (xhc-aln-x06.cisco.com [173.36.12.80]) by rcdn-core2-5.cisco.com (8.14.5/8.14.5) with ESMTP id r5RMmto4012025 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Thu, 27 Jun 2013 22:48:55 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.80]) by xhc-aln-x06.cisco.com ([173.36.12.80]) with mapi id 14.02.0318.004; Thu, 27 Jun 2013 17:48:55 -0500 From: "Jim Guichard (jguichar)" To: Ron Parker , "nsc@ietf.org" Thread-Topic: [nsc] Question regarding topology for bidirectional service chains Thread-Index: Ac5zaFiFaRP0gdxtSJ2ZgwSmZwBg6QAKIieA Date: Thu, 27 Jun 2013 22:48:55 +0000 Message-ID: <68B171751455884590F8E38E96416F363F55B755@xmb-rcd-x01.cisco.com> In-Reply-To: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.2.3.120616 x-originating-ip: [10.98.130.212] Content-Type: multipart/alternative; boundary="_000_68B171751455884590F8E38E96416F363F55B755xmbrcdx01ciscoc_" MIME-Version: 1.0 Subject: Re: [nsc] Question regarding topology for bidirectional service chains X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 27 Jun 2013 22:49:07 -0000 --_000_68B171751455884590F8E38E96416F363F55B755xmbrcdx01ciscoc_ Content-Type: text/plain; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable Hi Ron, I believe that both scenario's are valid; in some cases traffic will be req= uired to flow through both ingress and egress classifiers in both direction= s and in others it will not. In either case however a binding between east = and west chain must be established (so that the chain is in fact bidirectio= nal). From: Ron Parker > Date: Thursday, June 27, 2013 2:58 PM To: "nsc@ietf.org" > Subject: [nsc] Question regarding topology for bidirectional service chains I have a question for the group regarding bidirectional service chains. For sake of discussion, I will call the directions of packet flow East and = West. Let=92s also assume that classification-only nodes exist at each e= nd of the chain and that service-only nodes exist in the middle. And for = simplicity of discussion, let=92s say that service invocation order is an e= xact mirror in each direction. <---->classifier-West <---->service-node-1 <---->service-node-2 <---->= service-node-3 <---->classifier-East <----> In the eastbound direction, classifier-West first receives the traffic from= some upstream node and determines the service-chain. Likewise, in the we= stbound direction, classifier-East first receives the traffic from some ups= tream node and determines the service-chain. But, what about at the final= end of each chain? Must eastbound traffic traverse classifier-East and = must westbound traffic traverse classifier-West? I=92m suspecting that the answer is no, but it does create the asymmetrical= requirement that a non-classifier node can never be NLF-ingress, but can b= e NLF-egress. -->classifier-West -->service-node-1 -->service-node-2 -->service-node= -3 --> <--service-node-1 <--service-node-2 <--service-node-3 <--classifier-Ea= st <-- Thanks. Ron --_000_68B171751455884590F8E38E96416F363F55B755xmbrcdx01ciscoc_ Content-Type: text/html; charset="Windows-1252" Content-ID: <0F7491348E832F47B1B69F8A3079A2E1@emea.cisco.com> Content-Transfer-Encoding: quoted-printable
Hi Ron,

I believe that both scenario's are valid; in some cases traffic will b= e required to flow through both ingress and egress classifiers in both dire= ctions and in others it will not. In either case however a binding between = east and west chain must be established (so that the chain is in fact bidirectional). 

From: Ron Parker <Ron_Parker@affirmednetworks.com>= ;
Date: Thursday, June 27, 2013 2:58 = PM
To: "nsc@ietf.org" <nsc@ietf.= org>
Subject: [nsc] Question regarding t= opology for bidirectional service chains

I have a question for the group regarding bidirectio= nal service chains.  

 

For sake of discussion, I will call the directions o= f packet flow East and West.    Let=92s also assume that cla= ssification-only nodes exist at each end of the chain and that service-only= nodes exist in the middle.   And for simplicity of discussion, let=92s say that service invocation order is an exact mirror i= n each direction.

 

     =DF=E0classi= fier-West =DF=E0service-node-1 =DF=E0service-node-2 =DF=E0service-node-3 =DF=E0classifier-East =DF=E0

 

In the eastbound direction, classifier-West first re= ceives the traffic from some upstream node and determines the service-chain= .   Likewise, in the westbound direction, classifier-East first r= eceives the traffic from some upstream node and determines the service-chain.   But, what about at the final= end of each chain?    Must eastbound traffic traverse class= ifier-East and must westbound traffic traverse classifier-West?<= /p>

 

I=92m suspecting that the answer is no, but it does = create the asymmetrical requirement that a non-classifier node can never be= NLF-ingress, but can be NLF-egress.

 

     =E0classifier-West =E0service-node-1 =E0service-node-2 =E0se= rvice-node-3 =E0

     =DFservice-node-1 =DFservice-node-2 =DFservice-node-3 =DFcl= assifier-East =DF

 

 

Thanks.

 

   Ron

 --_000_68B171751455884590F8E38E96416F363F55B755xmbrcdx01ciscoc_-- From rraszuk@gmail.com Thu Jun 27 16:34:02 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 5971C11E812E for ; Thu, 27 Jun 2013 16:34:02 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -1.978 X-Spam-Level: X-Spam-Status: No, score=-1.978 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, FM_FORGED_GMAIL=0.622, NO_RELAYS=-0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id d2Bd8KcPiRkW for ; Thu, 27 Jun 2013 16:34:01 -0700 (PDT) Received: from mail-ie0-x235.google.com (mail-ie0-x235.google.com [IPv6:2607:f8b0:4001:c03::235]) by ietfa.amsl.com (Postfix) with ESMTP id C43AB11E812A for ; Thu, 27 Jun 2013 16:34:01 -0700 (PDT) Received: by mail-ie0-f181.google.com with SMTP id x12so2573098ief.26 for ; Thu, 27 Jun 2013 16:34:01 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:sender:in-reply-to:references:date :x-google-sender-auth:message-id:subject:from:to:cc:content-type :content-transfer-encoding; bh=mTK26wUMwEf51bfmUGaIi/ykhaY726E1UZN6F7NltMM=; b=MlDadUt0PskEdNs8HcEBprUgxRTG8JsDHM6OB0Hf+bSzKwAxSM5Ji41R61l9fGT0Lz Udve3ogiRBSB3T7gnt+QSNYTIWPiZCdOO3T+X5vPo1XqEeAphjaxdjiNfvtVcLYMzQTw 6daoyRQi/CAo6E/k8Q9aoziWaTB5p1icYZ7UtwIsNfRUW1IDiqTv/klCaG0xonZIepU7 hvF6bFgnqqMce4ZIC2kN4J11tHYPI1o/fRiGNG7iYi5qbHuYD7ujHpSJPr85xhIxAZcI 51H78iV/uvzk55rRfdr2MbH6Xzp/DRgHd7rt2/DjkCrkZ2hv0e1NhvDDYXiFEdwJ+Pw/ jdOw== MIME-Version: 1.0 X-Received: by 10.50.18.5 with SMTP id s5mr1156042igd.6.1372376041380; Thu, 27 Jun 2013 16:34:01 -0700 (PDT) Sender: rraszuk@gmail.com Received: by 10.64.28.168 with HTTP; Thu, 27 Jun 2013 16:34:01 -0700 (PDT) In-Reply-To: <68B171751455884590F8E38E96416F363F55B755@xmb-rcd-x01.cisco.com> References: <68B171751455884590F8E38E96416F363F55B755@xmb-rcd-x01.cisco.com> Date: Fri, 28 Jun 2013 01:34:01 +0200 X-Google-Sender-Auth: G2Sgr7t1Y9zouXZeTwJnAq7PzbU Message-ID: From: Robert Raszuk To: "Jim Guichard (jguichar)" Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: quoted-printable Cc: "nsc@ietf.org" , Ron Parker Subject: Re: [nsc] Question regarding topology for bidirectional service chains X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 27 Jun 2013 23:34:02 -0000 Hi Jim, I would agree with your conclusion. However bidirectional flows can be accomplished at the network level or at the network orchestration level. I think proceeding in both planes may unnecessarily complicate things. Do you really think bidirectional flows must be happening at the network level ? I think much easier could be to provide them at the higher levels. Regards, R. > On Fri, Jun 28, 2013 at 12:48 AM, Jim Guichard (jguichar) wrote: > > Hi Ron, > > I believe that both scenario's are valid; in some cases traffic will be > required to flow through both ingress and egress classifiers in both > directions and in others it will not. In either case however a binding > between east and west chain must be established (so that the chain is in > fact bidirectional). > > From: Ron Parker > Date: Thursday, June 27, 2013 2:58 PM > To: "nsc@ietf.org" > Subject: [nsc] Question regarding topology for bidirectional service chai= ns > > I have a question for the group regarding bidirectional service chains. > > > > For sake of discussion, I will call the directions of packet flow East an= d > West. Let=92s also assume that classification-only nodes exist at each= end > of the chain and that service-only nodes exist in the middle. And for > simplicity of discussion, let=92s say that service invocation order is an > exact mirror in each direction. > > > > =DF=E0classifier-West =DF=E0service-node-1 =DF=E0service-node-2 =DF= =E0service-node-3 > =DF=E0classifier-East =DF=E0 > > > > In the eastbound direction, classifier-West first receives the traffic fr= om > some upstream node and determines the service-chain. Likewise, in the > westbound direction, classifier-East first receives the traffic from some > upstream node and determines the service-chain. But, what about at the > final end of each chain? Must eastbound traffic traverse classifier-Ea= st > and must westbound traffic traverse classifier-West? > > > > I=92m suspecting that the answer is no, but it does create the asymmetric= al > requirement that a non-classifier node can never be NLF-ingress, but can = be > NLF-egress. > > > > =E0classifier-West =E0service-node-1 =E0service-node-2 =E0service-no= de-3 =E0 > > =DFservice-node-1 =DFservice-node-2 =DFservice-node-3 =DFclassifier-= East =DF > > > > > > Thanks. > > > > Ron > > > _______________________________________________ > nsc mailing list > nsc@ietf.org > https://www.ietf.org/mailman/listinfo/nsc > From jguichar@cisco.com Thu Jun 27 16:39:58 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 8092211E8104 for ; Thu, 27 Jun 2013 16:39:58 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -9.298 X-Spam-Level: X-Spam-Status: No, score=-9.298 tagged_above=-999 required=5 tests=[AWL=1.301, BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id XuwwScTPbOeD for ; Thu, 27 Jun 2013 16:39:53 -0700 (PDT) Received: from rcdn-iport-9.cisco.com (rcdn-iport-9.cisco.com [173.37.86.80]) by ietfa.amsl.com (Postfix) with ESMTP id A4F1911E810B for ; Thu, 27 Jun 2013 16:39:53 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=3092; q=dns/txt; s=iport; t=1372376393; x=1373585993; h=from:to:cc:subject:date:message-id:references: in-reply-to:content-transfer-encoding:mime-version; bh=45cQg/P8H91rU3mPFcmpudtzHAkZkTRUnHdJhYoh2Ao=; b=OkD4/7GjyFGCb9w+c3eWWsrqL9SwHdbu1+0NsF6ofaC98sAc/uOudyxy 3oR5PnD3+BNcCgJ7LmELEXJM0WUuDWqjazipjqZ6/azY6ZeXxQ5Tct+FL pyUvv48Bk7tDEHW2Gc+L54F6f/5VzJjlclbAhMlb12eMGay+feYqeD05H g=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AhsFAADMzFGtJXG8/2dsb2JhbABbgwkxv1WBAhZ0giMBAQEBAgEBAQFrCwULAgEIEQMBAgEuJwsdCAEBBA4FiAgGDLswBI8iMweDAmMDk3ODUoovhxaDEQ X-IronPort-AV: E=Sophos;i="4.87,955,1363132800"; d="scan'208";a="225408875" Received: from rcdn-core2-1.cisco.com ([173.37.113.188]) by rcdn-iport-9.cisco.com with ESMTP; 27 Jun 2013 23:39:53 +0000 Received: from xhc-aln-x05.cisco.com (xhc-aln-x05.cisco.com [173.36.12.79]) by rcdn-core2-1.cisco.com (8.14.5/8.14.5) with ESMTP id r5RNdrao004730 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Thu, 27 Jun 2013 23:39:53 GMT Received: from xmb-rcd-x01.cisco.com ([169.254.1.80]) by xhc-aln-x05.cisco.com ([173.36.12.79]) with mapi id 14.02.0318.004; Thu, 27 Jun 2013 18:39:52 -0500 From: "Jim Guichard (jguichar)" To: Robert Raszuk Thread-Topic: [nsc] Question regarding topology for bidirectional service chains Thread-Index: Ac5zaFiFaRP0gdxtSJ2ZgwSmZwBg6QAKIieAAAn1S4D//63RFg== Date: Thu, 27 Jun 2013 23:39:52 +0000 Message-ID: <74F89172-5388-4C7E-A589-CC1D3115CAE1@cisco.com> References: <68B171751455884590F8E38E96416F363F55B755@xmb-rcd-x01.cisco.com>, In-Reply-To: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: Content-Type: text/plain; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Cc: "nsc@ietf.org" , Ron Parker Subject: Re: [nsc] Question regarding topology for bidirectional service chains X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 27 Jun 2013 23:39:58 -0000 Hi Robert, I agree. Personally I think they should happen at the network orchestration= level and be pushed to the relevant network nodes. Sent from my iPhone On Jun 27, 2013, at 7:34 PM, "Robert Raszuk" wrote: > Hi Jim, >=20 > I would agree with your conclusion. >=20 > However bidirectional flows can be accomplished at the network level > or at the network orchestration level. >=20 > I think proceeding in both planes may unnecessarily complicate things. > Do you really think bidirectional flows must be happening at the > network level ? I think much easier could be to provide them at the > higher levels. >=20 > Regards, > R. >=20 >> On Fri, Jun 28, 2013 at 12:48 AM, Jim Guichard (jguichar) wrote: >>=20 >> Hi Ron, >>=20 >> I believe that both scenario's are valid; in some cases traffic will be >> required to flow through both ingress and egress classifiers in both >> directions and in others it will not. In either case however a binding >> between east and west chain must be established (so that the chain is in >> fact bidirectional). >>=20 >> From: Ron Parker >> Date: Thursday, June 27, 2013 2:58 PM >> To: "nsc@ietf.org" >> Subject: [nsc] Question regarding topology for bidirectional service cha= ins >>=20 >> I have a question for the group regarding bidirectional service chains. >>=20 >>=20 >>=20 >> For sake of discussion, I will call the directions of packet flow East a= nd >> West. Let=92s also assume that classification-only nodes exist at eac= h end >> of the chain and that service-only nodes exist in the middle. And for >> simplicity of discussion, let=92s say that service invocation order is a= n >> exact mirror in each direction. >>=20 >>=20 >>=20 >> =DF=E0classifier-West =DF=E0service-node-1 =DF=E0service-node-2 =DF= =E0service-node-3 >> =DF=E0classifier-East =DF=E0 >>=20 >>=20 >>=20 >> In the eastbound direction, classifier-West first receives the traffic f= rom >> some upstream node and determines the service-chain. Likewise, in the >> westbound direction, classifier-East first receives the traffic from som= e >> upstream node and determines the service-chain. But, what about at the >> final end of each chain? Must eastbound traffic traverse classifier-E= ast >> and must westbound traffic traverse classifier-West? >>=20 >>=20 >>=20 >> I=92m suspecting that the answer is no, but it does create the asymmetri= cal >> requirement that a non-classifier node can never be NLF-ingress, but can= be >> NLF-egress. >>=20 >>=20 >>=20 >> =E0classifier-West =E0service-node-1 =E0service-node-2 =E0service-no= de-3 =E0 >>=20 >> =DFservice-node-1 =DFservice-node-2 =DFservice-node-3 =DFclassifier-= East =DF >>=20 >>=20 >>=20 >>=20 >>=20 >> Thanks. >>=20 >>=20 >>=20 >> Ron >>=20 >>=20 >> _______________________________________________ >> nsc mailing list >> nsc@ietf.org >> https://www.ietf.org/mailman/listinfo/nsc >>=20 From rraszuk@gmail.com Thu Jun 27 16:41:33 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0EBF621F9CF1 for ; Thu, 27 Jun 2013 16:41:33 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -1.978 X-Spam-Level: X-Spam-Status: No, score=-1.978 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, FM_FORGED_GMAIL=0.622, NO_RELAYS=-0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id eZQ6THvVPE5G for ; Thu, 27 Jun 2013 16:41:32 -0700 (PDT) Received: from mail-ie0-x229.google.com (mail-ie0-x229.google.com [IPv6:2607:f8b0:4001:c03::229]) by ietfa.amsl.com (Postfix) with ESMTP id 9D7BA21F9CE9 for ; Thu, 27 Jun 2013 16:41:32 -0700 (PDT) Received: by mail-ie0-f169.google.com with SMTP id 10so2929977ied.0 for ; Thu, 27 Jun 2013 16:41:32 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:sender:in-reply-to:references:date :x-google-sender-auth:message-id:subject:from:to:cc:content-type; bh=H9E6m0eFGI1IoU7ta/9nEWxDEZ1tDkk2v8sTtMztYDw=; b=MkL3eR4JdoYtR+CAX0BG0a7R+OHF3fYyne1/uLI+4lvHKOyWy9r15fphh7jtu/GE6/ CypKY3D7GZVcUm2lS6ZQQ661Lkjt06jshlw+L+3oRpskRkNQ6FH4yLIIzkj4bQ9E7g5W 0SakQClz4RHUDLhLw0A+9rlF9X0Mh9yr/bugCA1am24XNqYQI+fvrpuyd3VW+1oYalK/ 1o8Uhum6MacJVVuzLj86vruXRgmyJToF+4Ff7FZvse/Mehz8Sb1nWcMU6RFwF5wAR2Gw BUNdRkFpNc/NTkJ/gsGoWYymGsq36j19E5QeKB2gU4Qmm3Xx6r0SZkCNCoEq0U6Sh0JH jT4w== MIME-Version: 1.0 X-Received: by 10.50.107.65 with SMTP id ha1mr975565igb.50.1372376492254; Thu, 27 Jun 2013 16:41:32 -0700 (PDT) Sender: rraszuk@gmail.com Received: by 10.64.28.168 with HTTP; Thu, 27 Jun 2013 16:41:32 -0700 (PDT) In-Reply-To: <74F89172-5388-4C7E-A589-CC1D3115CAE1@cisco.com> References: <68B171751455884590F8E38E96416F363F55B755@xmb-rcd-x01.cisco.com> <74F89172-5388-4C7E-A589-CC1D3115CAE1@cisco.com> Date: Fri, 28 Jun 2013 01:41:32 +0200 X-Google-Sender-Auth: BOQuoUAUiQzfVr9adfFryAMH7kk Message-ID: From: Robert Raszuk To: "Jim Guichard (jguichar)" Content-Type: text/plain; charset=ISO-8859-1 Cc: "nsc@ietf.org" , Ron Parker Subject: Re: [nsc] Question regarding topology for bidirectional service chains X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 27 Jun 2013 23:41:33 -0000 Same here ! Great, R. On Fri, Jun 28, 2013 at 1:39 AM, Jim Guichard (jguichar) wrote: > Hi Robert, > > I agree. Personally I think they should happen at the network orchestration level and be pushed to the relevant network nodes. From liushucheng@huawei.com Thu Jun 27 19:29:49 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 80F5921F9B53 for ; Thu, 27 Jun 2013 19:29:49 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id qGrCCeuD+sTR for ; Thu, 27 Jun 2013 19:29:45 -0700 (PDT) Received: from lhrrgout.huawei.com (lhrrgout.huawei.com [194.213.3.17]) by ietfa.amsl.com (Postfix) with ESMTP id EB67121F9B4C for ; Thu, 27 Jun 2013 19:29:44 -0700 (PDT) Received: from 172.18.7.190 (EHLO lhreml203-edg.china.huawei.com) ([172.18.7.190]) by lhrrg01-dlp.huawei.com (MOS 4.3.5-GA FastPath queued) with ESMTP id AUL91865; Fri, 28 Jun 2013 02:29:43 +0000 (GMT) Received: from LHREML404-HUB.china.huawei.com (10.201.5.218) by lhreml203-edg.huawei.com (172.18.7.221) with Microsoft SMTP Server (TLS) id 14.1.323.7; Fri, 28 Jun 2013 03:28:44 +0100 Received: from SZXEML451-HUB.china.huawei.com (10.82.67.194) by lhreml404-hub.china.huawei.com (10.201.5.218) with Microsoft SMTP Server (TLS) id 14.1.323.7; Fri, 28 Jun 2013 10:29:41 +0800 Received: from SZXEML546-MBX.china.huawei.com ([169.254.3.95]) by szxeml451-hub.china.huawei.com ([10.82.67.194]) with mapi id 14.01.0323.007; Fri, 28 Jun 2013 10:29:27 +0800 From: "Will Liu (Shucheng)" To: "nsc@ietf.org" Thread-Topic: New Version Notification for draft-liu-service-chaining-use-cases-00.txt Thread-Index: AQHOc0cwbvJY5kJVJ0GPLe64ILUHCJlKZmug Date: Fri, 28 Jun 2013 02:29:26 +0000 Message-ID: Accept-Language: en-US, zh-CN Content-Language: zh-CN X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [10.66.78.79] Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: base64 MIME-Version: 1.0 X-CFilter-Loop: Reflected Cc: "Hongyu Li \(Julio\)" Subject: [nsc] FW: New Version Notification for draft-liu-service-chaining-use-cases-00.txt X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 28 Jun 2013 02:29:49 -0000 SGkgRm9sa3MsDQoNCldlIGhhdmUgc3VibWl0dGVkIGEgZHJhZnQgZGlzY3Vzc2luZyBhYm91dCB0 aGUgdXNlIGNhc2Ugb2Ygc2VydmljZSBjaGFpbmluZyBhcyBiZWxvdy4gVGhpcyBtYXkgaGVscCBw ZW9wbGUgdG8gdW5kZXJzdGFuZCB0aGUgdXNlIG9mIHNlcnZpY2UgY2hhaW4uIFlvdXIgY29tbWVu dHMgYXJlIGhpZ2hseSBhcHByZWNpYXRlZC4NCg0KQ2hlZXJzLA0KV2lsbA0KDQoNCi0tLS0tT3Jp Z2luYWwgTWVzc2FnZS0tLS0tDQpGcm9tOiBpbnRlcm5ldC1kcmFmdHNAaWV0Zi5vcmcgW21haWx0 bzppbnRlcm5ldC1kcmFmdHNAaWV0Zi5vcmddIA0KU2VudDogVGh1cnNkYXksIEp1bmUgMjcsIDIw MTMgMTE6MDEgUE0NClRvOiBXaWxsIExpdSAoU2h1Y2hlbmcpOyBIdWFuZ3lvbmcgKE9saXZlcik7 IEhvbmd5dSBMaSAoSnVsaW8pOyBXaWxsIExpdSAoU2h1Y2hlbmcpDQpTdWJqZWN0OiBOZXcgVmVy c2lvbiBOb3RpZmljYXRpb24gZm9yIGRyYWZ0LWxpdS1zZXJ2aWNlLWNoYWluaW5nLXVzZS1jYXNl cy0wMC50eHQNCg0KDQpBIG5ldyB2ZXJzaW9uIG9mIEktRCwgZHJhZnQtbGl1LXNlcnZpY2UtY2hh aW5pbmctdXNlLWNhc2VzLTAwLnR4dA0KaGFzIGJlZW4gc3VjY2Vzc2Z1bGx5IHN1Ym1pdHRlZCBi eSBXaWxsKFNodWNoZW5nKSBMaXUgYW5kIHBvc3RlZCB0byB0aGUNCklFVEYgcmVwb3NpdG9yeS4N Cg0KRmlsZW5hbWU6CSBkcmFmdC1saXUtc2VydmljZS1jaGFpbmluZy11c2UtY2FzZXMNClJldmlz aW9uOgkgMDANClRpdGxlOgkJIFNlcnZpY2UgQ2hhaW5pbmcgVXNlIENhc2VzDQpDcmVhdGlvbiBk YXRlOgkgMjAxMy0wNi0yNw0KR3JvdXA6CQkgSW5kaXZpZHVhbCBTdWJtaXNzaW9uDQpOdW1iZXIg b2YgcGFnZXM6IDcNClVSTDogICAgICAgICAgICAgaHR0cDovL3d3dy5pZXRmLm9yZy9pbnRlcm5l dC1kcmFmdHMvZHJhZnQtbGl1LXNlcnZpY2UtY2hhaW5pbmctdXNlLWNhc2VzLTAwLnR4dA0KU3Rh dHVzOiAgICAgICAgICBodHRwOi8vZGF0YXRyYWNrZXIuaWV0Zi5vcmcvZG9jL2RyYWZ0LWxpdS1z ZXJ2aWNlLWNoYWluaW5nLXVzZS1jYXNlcw0KSHRtbGl6ZWQ6ICAgICAgICBodHRwOi8vdG9vbHMu aWV0Zi5vcmcvaHRtbC9kcmFmdC1saXUtc2VydmljZS1jaGFpbmluZy11c2UtY2FzZXMtMDANCg0K DQpBYnN0cmFjdDoNCiAgIEZvciBzb21lIG5ldHdvcmsgc2VydmljZXMsIHRyYWZmaWMgaXMgZm9y d2FyZGVkIHRocm91Z2ggYSBzZXF1ZW5jZSBvZg0KICAgbmV0d29yayBmdW5jdGlvbnMsIHdoaWNo IHVzdWFsbHkgaGF2ZSBkZWRpY2F0ZWQgY2FwYWJpbGl0aWVzIG90aGVyDQogICB0aGFuIGZvcndh cmRpbmcsIGUuZy4gZmlyZXdhbGwuICBGb3J3YXJkaW5nIG9mIHRyYWZmaWMgYWxvbmcgYQ0KICAg c2VxdWVuY2Ugb2Ygc2VydmljZSBwcm9jZXNzaW5nIGZ1bmN0aW9ucyBpcyB1c3VhbGx5IGJhc2Vk IG9uIHNlcnZpY2UNCiAgIGNoYXJhY3RlcmlzdGljcywgYnV0IG5vdCBkZXN0aW5hdGlvbiBhZGRy ZXNzZXMuICBXZSBjYWxsIHN1Y2ggd2F5IG9mDQogICBmb3J3YXJkaW5nIGFzIHNlcnZpY2UgY2hh aW5pbmcuICBUaGlzIG1lbW8gZGVzY3JpYmVzIHVzZSBjYXNlcyBvZg0KICAgc2VydmljZSBjaGFp bmluZy4NCg0KICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIA0KDQoNClRoZSBJRVRGIFNlY3JldGFy aWF0DQoNCg== From hongyu.li@huawei.com Fri Jun 28 01:00:59 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id B351621F9E27 for ; Fri, 28 Jun 2013 01:00:59 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -4 X-Spam-Level: X-Spam-Status: No, score=-4 tagged_above=-999 required=5 tests=[RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id cThVVcado19V for ; Fri, 28 Jun 2013 01:00:50 -0700 (PDT) Received: from lhrrgout.huawei.com (lhrrgout.huawei.com [194.213.3.17]) by ietfa.amsl.com (Postfix) with ESMTP id 9AFF221F9EBE for ; Fri, 28 Jun 2013 01:00:31 -0700 (PDT) Received: from 172.18.7.190 (EHLO lhreml203-edg.china.huawei.com) ([172.18.7.190]) by lhrrg01-dlp.huawei.com (MOS 4.3.5-GA FastPath queued) with ESMTP id AUM13804; Fri, 28 Jun 2013 08:00:28 +0000 (GMT) Received: from LHREML401-HUB.china.huawei.com (10.201.5.240) by lhreml203-edg.huawei.com (172.18.7.221) with Microsoft SMTP Server (TLS) id 14.1.323.7; Fri, 28 Jun 2013 08:59:26 +0100 Received: from SZXEML454-HUB.china.huawei.com (10.82.67.197) by lhreml401-hub.china.huawei.com (10.201.5.240) with Microsoft SMTP Server (TLS) id 14.1.323.7; Fri, 28 Jun 2013 09:00:24 +0100 Received: from szxeml560-mbx.china.huawei.com ([169.254.3.63]) by SZXEML454-HUB.china.huawei.com ([10.82.67.197]) with mapi id 14.01.0323.007; Fri, 28 Jun 2013 16:00:18 +0800 From: "Hongyu Li (Julio)" To: "nsc@ietf.org" Thread-Topic: New Version Notification for draft-li-service-chaining-requirements-00.txt Thread-Index: AQHOcz6FW8l/t7Ua20mtoZGWKgoa85lKxAJA Date: Fri, 28 Jun 2013 08:00:17 +0000 Message-ID: <6EB34CB5D82C4645B826C56144826EA97DF0B859@szxeml560-mbx.china.huawei.com> Accept-Language: en-US, zh-CN Content-Language: zh-CN X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [10.66.114.234] Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: base64 MIME-Version: 1.0 X-CFilter-Loop: Reflected Subject: [nsc] FW: New Version Notification for draft-li-service-chaining-requirements-00.txt X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 28 Jun 2013 08:00:59 -0000 SGksDQoNCldlIGhhdmUgdXBsb2FkZWQgYSBkcmFmdCBvbiBOZXR3b3JrIFNlcnZpY2UgQ2hhaW5p bmcgcmVxdWlyZW1lbnRzLiBObyBzb2x1dGlvbiBwcm9wb3NlZC4NCg0KQ29tbWVudHMgYXJlIHdl bGNvbWUuDQoNCkNoZWVycywNCkhvbmd5dQ0KDQotLS0tLU9yaWdpbmFsIE1lc3NhZ2UtLS0tLQ0K RnJvbTogaW50ZXJuZXQtZHJhZnRzQGlldGYub3JnIFttYWlsdG86aW50ZXJuZXQtZHJhZnRzQGll dGYub3JnXSANClNlbnQ6IFRodXJzZGF5LCBKdW5lIDI3LCAyMDEzIDk6NTkgUE0NClRvOiBIb25n eXUgTGkgKEp1bGlvKQ0KU3ViamVjdDogTmV3IFZlcnNpb24gTm90aWZpY2F0aW9uIGZvciBkcmFm dC1saS1zZXJ2aWNlLWNoYWluaW5nLXJlcXVpcmVtZW50cy0wMC50eHQNCg0KDQpBIG5ldyB2ZXJz aW9uIG9mIEktRCwgZHJhZnQtbGktc2VydmljZS1jaGFpbmluZy1yZXF1aXJlbWVudHMtMDAudHh0 DQpoYXMgYmVlbiBzdWNjZXNzZnVsbHkgc3VibWl0dGVkIGJ5IEhvbmd5dSBMaSBhbmQgcG9zdGVk IHRvIHRoZQ0KSUVURiByZXBvc2l0b3J5Lg0KDQpGaWxlbmFtZToJIGRyYWZ0LWxpLXNlcnZpY2Ut Y2hhaW5pbmctcmVxdWlyZW1lbnRzDQpSZXZpc2lvbjoJIDAwDQpUaXRsZToJCSBSZXF1aXJlbWVu dHMgZm9yIFNlcnZpY2UgQ2hhaW5pbmcNCkNyZWF0aW9uIGRhdGU6CSAyMDEzLTA2LTI3DQpHcm91 cDoJCSBJbmRpdmlkdWFsIFN1Ym1pc3Npb24NCk51bWJlciBvZiBwYWdlczogNQ0KVVJMOiAgICAg ICAgICAgICBodHRwOi8vd3d3LmlldGYub3JnL2ludGVybmV0LWRyYWZ0cy9kcmFmdC1saS1zZXJ2 aWNlLWNoYWluaW5nLXJlcXVpcmVtZW50cy0wMC50eHQNClN0YXR1czogICAgICAgICAgaHR0cDov L2RhdGF0cmFja2VyLmlldGYub3JnL2RvYy9kcmFmdC1saS1zZXJ2aWNlLWNoYWluaW5nLXJlcXVp cmVtZW50cw0KSHRtbGl6ZWQ6ICAgICAgICBodHRwOi8vdG9vbHMuaWV0Zi5vcmcvaHRtbC9kcmFm dC1saS1zZXJ2aWNlLWNoYWluaW5nLXJlcXVpcmVtZW50cy0wMA0KDQoNCkFic3RyYWN0Og0KICAg Q2hhaW5pbmcgc29tZSBzZXJ2aWNlIHByb2Nlc3NpbmcgZnVuY3Rpb25zIG1heSBwcm92aWRlIHNl cnZpY2VzIGluIGENCiAgIGZsZXhpYmxlIHdheS4gIFN1Y2ggYSBtZXRob2QgaXMgY2FsbGVkIHNl cnZpY2UgY2hhaW5pbmcuICBUaGlzDQogICBkb2N1bWVudCBwcmVzZW50cyBhIGxpc3Qgb2YgZ2Vu ZXJhbCByZXF1aXJlbWVudHMgc2VydmljZSBjaGFpbmluZw0KICAgc2hvdWxkIHN1cHBvcnQuICBT b2x1dGlvbnMgb2Ygc2VydmljZSBjaGFpbmluZyBhcmUgbm90IHByZXNlbnRlZCBpbg0KICAgdGhp cyBkb2N1bWVudC4NCg0KICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIA0KDQoNClRoZSBJRVRGIFNl Y3JldGFyaWF0DQoNCg== From hongyu.li@huawei.com Fri Jun 28 01:01:58 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 2F63B21F9EE2 for ; Fri, 28 Jun 2013 01:01:58 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -4 X-Spam-Level: X-Spam-Status: No, score=-4 tagged_above=-999 required=5 tests=[RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id tDRvqh4-CLY8 for ; Fri, 28 Jun 2013 01:01:48 -0700 (PDT) Received: from lhrrgout.huawei.com (lhrrgout.huawei.com [194.213.3.17]) by ietfa.amsl.com (Postfix) with ESMTP id 21E9421F9EE1 for ; Fri, 28 Jun 2013 01:01:42 -0700 (PDT) Received: from 172.18.7.190 (EHLO lhreml203-edg.china.huawei.com) ([172.18.7.190]) by lhrrg01-dlp.huawei.com (MOS 4.3.5-GA FastPath queued) with ESMTP id AUM13946; Fri, 28 Jun 2013 08:01:42 +0000 (GMT) Received: from LHREML403-HUB.china.huawei.com (10.201.5.217) by lhreml203-edg.huawei.com (172.18.7.221) with Microsoft SMTP Server (TLS) id 14.1.323.7; Fri, 28 Jun 2013 09:00:42 +0100 Received: from SZXEML417-HUB.china.huawei.com (10.82.67.156) by lhreml403-hub.china.huawei.com (10.201.5.217) with Microsoft SMTP Server (TLS) id 14.1.323.7; Fri, 28 Jun 2013 09:01:40 +0100 Received: from szxeml560-mbx.china.huawei.com ([169.254.3.63]) by szxeml417-hub.china.huawei.com ([10.82.67.156]) with mapi id 14.01.0323.007; Fri, 28 Jun 2013 16:01:32 +0800 From: "Hongyu Li (Julio)" To: "nsc@ietf.org" Thread-Topic: New Version Notification for draft-jiang-service-chaining-arch-00.txt Thread-Index: AQHOcytisN8j4U9Hqku3RyRP3qseUplKw/Zg Date: Fri, 28 Jun 2013 08:01:31 +0000 Message-ID: <6EB34CB5D82C4645B826C56144826EA97DF0B865@szxeml560-mbx.china.huawei.com> Accept-Language: en-US, zh-CN Content-Language: zh-CN X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [10.66.114.234] Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: base64 MIME-Version: 1.0 X-CFilter-Loop: Reflected Subject: [nsc] FW: New Version Notification for draft-jiang-service-chaining-arch-00.txt X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 28 Jun 2013 08:01:58 -0000 SGksDQoNCldlIGhhdmUgdXBsb2FkZWQgYSBkcmFmdCBvbiBTZXJ2aWNlIENoYWluaW5nIEFyY2hp dGVjdHVyZS4gSXQgaW5jbHVkZXMgYSBnZW5lcmFsIGFyY2hpdGVjdHVyZSBhbmQgc29tZSBwb3Rl bnRpYWwgc29sdXRpb24gYW5hbHlzaXMuDQoNCkNvbW1lbnRzIGFyZSB3ZWxjb21lLg0KDQpDaGVl cnMsDQpIb25neXUNCg0KLS0tLS1PcmlnaW5hbCBNZXNzYWdlLS0tLS0NCkZyb206IGludGVybmV0 LWRyYWZ0c0BpZXRmLm9yZyBbbWFpbHRvOmludGVybmV0LWRyYWZ0c0BpZXRmLm9yZ10gDQpTZW50 OiBUaHVyc2RheSwgSnVuZSAyNywgMjAxMyA3OjQyIFBNDQpUbzogSG9uZ3l1IExpIChKdWxpbyk7 IEppYW5neXVhbmxvbmc7IEhvbmd5dSBMaSAoSnVsaW8pDQpTdWJqZWN0OiBOZXcgVmVyc2lvbiBO b3RpZmljYXRpb24gZm9yIGRyYWZ0LWppYW5nLXNlcnZpY2UtY2hhaW5pbmctYXJjaC0wMC50eHQN Cg0KDQpBIG5ldyB2ZXJzaW9uIG9mIEktRCwgZHJhZnQtamlhbmctc2VydmljZS1jaGFpbmluZy1h cmNoLTAwLnR4dA0KaGFzIGJlZW4gc3VjY2Vzc2Z1bGx5IHN1Ym1pdHRlZCBieSBZdWFubG9uZyBK aWFuZyBhbmQgcG9zdGVkIHRvIHRoZQ0KSUVURiByZXBvc2l0b3J5Lg0KDQpGaWxlbmFtZToJIGRy YWZ0LWppYW5nLXNlcnZpY2UtY2hhaW5pbmctYXJjaA0KUmV2aXNpb246CSAwMA0KVGl0bGU6CQkg QW4gQXJjaGl0ZWN0dXJlIG9mIFNlcnZpY2UgQ2hhaW5pbmcNCkNyZWF0aW9uIGRhdGU6CSAyMDEz LTA2LTI3DQpHcm91cDoJCSBJbmRpdmlkdWFsIFN1Ym1pc3Npb24NCk51bWJlciBvZiBwYWdlczog MTENClVSTDogICAgICAgICAgICAgaHR0cDovL3d3dy5pZXRmLm9yZy9pbnRlcm5ldC1kcmFmdHMv ZHJhZnQtamlhbmctc2VydmljZS1jaGFpbmluZy1hcmNoLTAwLnR4dA0KU3RhdHVzOiAgICAgICAg ICBodHRwOi8vZGF0YXRyYWNrZXIuaWV0Zi5vcmcvZG9jL2RyYWZ0LWppYW5nLXNlcnZpY2UtY2hh aW5pbmctYXJjaA0KSHRtbGl6ZWQ6ICAgICAgICBodHRwOi8vdG9vbHMuaWV0Zi5vcmcvaHRtbC9k cmFmdC1qaWFuZy1zZXJ2aWNlLWNoYWluaW5nLWFyY2gtMDANCg0KDQpBYnN0cmFjdDoNCiAgIEFz IG5ldHdvcmsgdmlydHVhbGl6YXRpb24gaXMgb3BlbmluZyB0aGUgZ2F0ZSB0byBtdWNoIG1vcmUg aW5ub3ZhdGl2ZQ0KICAgc2VydmljZXMgZm9yIHNlcnZpY2UgcHJvdmlkZXJzLCBzZXJ2aWNlIGNo YWluaW5nIHByb3ZpZGVzIGEgZmxleGlibGUNCiAgIHdheSBvZiBzZXJ2aWNlIHByb3Zpc2lvbmlu ZyBhbmQgZmFjaWxpdGF0ZXMgdGhlaXIgZGVwbG95bWVudC4NCg0KICAgVGhpcyBkb2N1bWVudCBw cm92aWRlcyBhIGdlbmVyYWwgYWJzdHJhY3QgYXJjaGl0ZWN0dXJlIGZvciBzZXJ2aWNlDQogICBj aGFpbmluZy4gSXQgaXMgYSBmbGV4aWJsZSBhbmQgc2NhbGFibGUgYXJjaGl0ZWN0dXJlIHdoaWNo IGNhbg0KICAgZnVsZmlsbCByZXF1aXJlbWVudHMgb2Ygc2VydmljZSBjaGFpbmluZy4gU29tZSBz b2x1dGlvbnMgYmFzZWQgb24NCiAgIHRoaXMgYXJjaGl0ZWN0dXJlIGFyZSBhbHNvIGRpc2N1c3Nl ZCB3aXRoIHRoZWlyIGFkdmFudGFnZXMgYW5kDQogICBkaXNhZHZhbnRhZ2VzLiBUaGlzIGFyY2hp dGVjdHVyZSBjYW4gYmUgdXNlZCBhcyBhIGd1aWRlbGluZSBhbmQgYWxzbw0KICAgYSBjcml0ZXJp b24gZm9yIHRoZSBkZXNpZ24gb2Ygc2VydmljZSBjaGFpbmluZy4NCg0KICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgIA0KDQoNClRoZSBJRVRGIFNlY3JldGFyaWF0DQoNCg== From mohamed.boucadair@orange.com Fri Jun 28 05:24:23 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D301521F9D2D for ; Fri, 28 Jun 2013 05:24:23 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.248 X-Spam-Level: X-Spam-Status: No, score=-2.248 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HELO_EQ_FR=0.35, UNPARSEABLE_RELAY=0.001] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id T2SH6NItIg9d for ; Fri, 28 Jun 2013 05:24:14 -0700 (PDT) Received: from relais-inet.francetelecom.com (relais-ias91.francetelecom.com [193.251.215.91]) by ietfa.amsl.com (Postfix) with ESMTP id DF2C421F9D6D for ; Fri, 28 Jun 2013 05:23:53 -0700 (PDT) Received: from omfedm08.si.francetelecom.fr (unknown [xx.xx.xx.4]) by omfedm11.si.francetelecom.fr (ESMTP service) with ESMTP id 93F503B48D9 for ; Fri, 28 Jun 2013 14:23:49 +0200 (CEST) Received: from PUEXCH21.nanterre.francetelecom.fr (unknown [10.101.44.28]) by omfedm08.si.francetelecom.fr (ESMTP service) with ESMTP id 78F86238048 for ; Fri, 28 Jun 2013 14:23:49 +0200 (CEST) Received: from PUEXCB1B.nanterre.francetelecom.fr ([10.101.44.13]) by PUEXCH21.nanterre.francetelecom.fr ([10.101.44.28]) with mapi; Fri, 28 Jun 2013 14:23:49 +0200 From: To: "nsc@ietf.org" Date: Fri, 28 Jun 2013 14:23:48 +0200 Thread-Topic: draft-boucadair-network-function-chaining-01 Thread-Index: Ac5z+lcbTULkEArHROCzVXpNygG6IQ== Message-ID: <94C682931C08B048B7A8645303FDC9F36EDB6BB361@PUEXCB1B.nanterre.francetelecom.fr> Accept-Language: fr-FR Content-Language: fr-FR X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: fr-FR Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-PMX-Version: 5.6.1.2065439, Antispam-Engine: 2.7.2.376379, Antispam-Data: 2013.6.28.101520 Subject: [nsc] draft-boucadair-network-function-chaining-01 X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 28 Jun 2013 12:24:23 -0000 Dear all, A new version integrating the comments we received is available online. Suggestions, contributions, questions are more than welcome. Cheers, Med -----Message d'origine----- De=A0: i-d-announce-bounces@ietf.org [mailto:i-d-announce-bounces@ietf.org]= De la part de internet-drafts@ietf.org Envoy=E9=A0: vendredi 28 juin 2013 14:11 =C0=A0: i-d-announce@ietf.org Objet=A0: I-D Action: draft-boucadair-network-function-chaining-01.txt A New Internet-Draft is available from the on-line Internet-Drafts director= ies. Title : Differentiated Network-Located Function Chaining Framewo= rk Author(s) : Mohamed Boucadair Christian Jacquenet Ron Parker Diego R. Lopez Parviz Yegani Filename : draft-boucadair-network-function-chaining-01.txt Pages : 19 Date : 2013-06-28 Abstract: IP networks rely more and more on the combination of advanced functions (besides the basic routing and forwarding functions). This document defines a solution to enforce Network-Located Function Chaining (NLFC) with minimum requirements on the underlying network. The proposed solution allows for Differentiated Forwarding (DiffForward): packets are classified at the entry point of an NLFC- enabled network, and are then forwarded on a per NLFC map basis. The IETF datatracker status page for this draft is: https://datatracker.ietf.org/doc/draft-boucadair-network-function-chaining There's also a htmlized version available at: http://tools.ietf.org/html/draft-boucadair-network-function-chaining-01 A diff from the previous version is available at: http://www.ietf.org/rfcdiff?url2=3Ddraft-boucadair-network-function-chainin= g-01 Internet-Drafts are also available by anonymous FTP at: ftp://ftp.ietf.org/internet-drafts/ _______________________________________________ I-D-Announce mailing list I-D-Announce@ietf.org https://www.ietf.org/mailman/listinfo/i-d-announce Internet-Draft directories: http://www.ietf.org/shadow.html or ftp://ftp.ietf.org/ietf/1shadow-sites.txt From smkumar@cisco.com Fri Jun 28 11:58:34 2013 Return-Path: X-Original-To: nsc@ietfa.amsl.com Delivered-To: nsc@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7EFA221F854E for ; Fri, 28 Jun 2013 11:58:34 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -10.599 X-Spam-Level: X-Spam-Status: No, score=-10.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8] Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 4vG3f8SYfKNj for ; Fri, 28 Jun 2013 11:58:29 -0700 (PDT) Received: from rcdn-iport-8.cisco.com (rcdn-iport-8.cisco.com [173.37.86.79]) by ietfa.amsl.com (Postfix) with ESMTP id 22F0B21F85BB for ; Fri, 28 Jun 2013 11:58:25 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=11583; q=dns/txt; s=iport; t=1372445905; x=1373655505; h=from:to:cc:subject:date:message-id:in-reply-to: content-id:content-transfer-encoding:mime-version; bh=ZUkHWDq6AWX3I1vqckZUNZ+l5Qu/O9Xf6qAjAWWyEpo=; b=LEYiS3x5fpEeQxSQfg/bzhIKElQxqch/FIWoAbas6ElNTP3xU2F/gloc ThcuJqWz0RJqNW5rIoR+QPJrz0XgtMMkg2oXmDy78MJ0V+LNkDkDw2sDJ VZfKAKrDK2OeJ/9iohq7pOOUtqTdAll+D1FssOGNvOAYx6YHuatbrsc1z s=; X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AhQFABDbzVGtJV2c/2dsb2JhbABRAQmDCTJJvwaBChZ0giMBAQEEAQEBCywrAgcLDAYBCBEEAQELFAkuCxQJCAEBBAENBQiIBwy6PASOGgGBCgYrBwaCfmMDqQqDEYFqJBo X-IronPort-AV: E=Sophos;i="4.87,960,1363132800"; d="scan'208";a="228759316" Received: from rcdn-core-5.cisco.com ([173.37.93.156]) by rcdn-iport-8.cisco.com with ESMTP; 28 Jun 2013 18:58:24 +0000 Received: from xhc-aln-x10.cisco.com (xhc-aln-x10.cisco.com [173.36.12.84]) by rcdn-core-5.cisco.com (8.14.5/8.14.5) with ESMTP id r5SIwOLO007339 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=FAIL); Fri, 28 Jun 2013 18:58:24 GMT Received: from xmb-aln-x09.cisco.com ([169.254.4.92]) by xhc-aln-x10.cisco.com ([173.36.12.84]) with mapi id 14.02.0318.004; Fri, 28 Jun 2013 13:58:22 -0500 From: "Surendra Kumar (smkumar)" To: Ron Parker , "Paul Quinn (paulq)" Thread-Topic: [nsc] Comments on draft-quinn-nsh-00 Thread-Index: Ac5w7gaFkR09Y/lqTveMLRqlAXe6pQBvkqEAAACgfYAAXHeJAA== Date: Fri, 28 Jun 2013 18:58:22 +0000 Message-ID: <8E86B6E7C6FB9F40A00B0E269045A7C91DDBB551@xmb-aln-x09.cisco.com> In-Reply-To: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: user-agent: Microsoft-MacOutlook/14.3.5.130515 x-originating-ip: [10.21.80.216] Content-Type: text/plain; charset="us-ascii" Content-ID: Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Cc: "nsc@ietf.org" Subject: Re: [nsc] Comments on draft-quinn-nsh-00 X-BeenThere: nsc@ietf.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Network Service Chaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 28 Jun 2013 18:58:34 -0000 Ron, See my comments inline - SK> Rgds, Surendra. On 6/26/13 5:50 PM, "Ron Parker" wrote: >Hi, Paul. > >Thanks for the response. My follow up embedded. > >BR, >Ron > >-----Original Message----- >From: Paul Quinn [mailto:paulq@cisco.com] >Sent: Wednesday, June 26, 2013 11:33 AM >To: Ron Parker >Cc: nsc@ietf.org >Subject: Re: [nsc] Comments on draft-quinn-nsh-00 > >Hi Ron, > >Thanks for the thoughtful comments! My apologies for not getting back to >you sooner, I've been traveling. > >See my comments inline. > >Paul > >On Jun 24, 2013, at 8:19 AM, Ron Parker >wrote: > >> I am new to this group and want to thank the authors and participants >>for the good work you've done so far. I feel that this work is very >>complementary to the Network Function Virtualization Forum's nascent >>network function chaining work. Draft-quinn-nsh-00 and >>draft-quinn-nsc-problem-statement-00 are, however, appropriately more >>general in that mobile data services are not assumed to be the de facto >>business problem and co-location of network functions are not assumed to >>be within the data center. >>=20 > >PQ> Yes. We make no assumption on applicability to particular market >segments or location of service functions - all are within scope of this >work.=20 > > > >> My specific questions and comments on draft-quinn-nsh-00 follow... >>=20 >> General - I agree with some previous comments on the thread that >>"service" is an overloaded term. I prefer the NFV's terms "network >>function" and "network function chaining". Let the business folks >>determine what set of functions constitute a "service" that consumers or >>businesses will pay for. > >PQ> We tried use the term service function to describe the specific >"services": e.g. firewall. As Jim Guichard mentioned on the alias: >services are what a customer buys and those services could be constructed >using one or more "service functions". I don't want to use the term >network function as that implies that the network somehow does something >here which is not the case as the function applied is not necessarily at >the network layer. >Ron> Makes sense. > > >>=20 >> 2.2 Problem Statement; 1. Topological Dependencies - This would be a >>good place to clarify whether service chains are unidirectional or >>bidirectional. If bidirectional, is symmetry assumed (i.e. if from >>client to server network functions A, B, C are invoked, then from server >>to client, is it always C, B, A? Must a classifier be at both ends of >>the chain? Must it be the same classifier at each end? >>=20 > >PQ> Point well taken, we have two choices: do you have suggested text >you think captures this, or I'm happy to send out a revision that >incorporates comments to date early next week. One thing to note: >service chains could be unidirectional, bidirectional or something >in-between and that symmetry may or may not be required. As part of the >work we will have to define the "types" of service chains and this is >something that the WG would be assigned to do. >Ron> I don't have specific text at this point. Agree that >unidirectional, inversely-symmetrical bidirectional, and asymmetrical >bidirectional are all valid. Suggest that the flexible location of the >classifier(s) be brought out, too. > > > > >> 2.2 Problem Statement - It may be useful to discuss how standard IP >>routing and NSC co-exist. For example, a deployment based on this >>technique may utilize IP routing to ensure packets arrive at the two >>ends of a service chain, but the service overlay is used to steer >>traffic between the two ends of the chain. >>=20 > >PQ> Can you elaborate a bit more? In general, routing carries the >encapsulated packets from service to service. The chaining function >provides the service function destination for routing lookup. IP >routing is simply used for reachability of each service function within a >service chain.=20 >Ron> I feel that many people thinking about this topic have somehow >arrived at a conclusion that OpenFlow is automatically and exclusively >the basis for achieving things like NSC. SK> That is a fallacy! What is needed is a mechanism to steer the traffic (frames/packets) to one or more service nodes while being able to identify the path used to steer the traffic. A simple overly network in the "data plane" with a service header like NSH satisfies this. > > >> 3.1 NSH Actions; 1. Insert / Remove Service Header - this section >>introduces the concept of control plane to indicate which node is "end >>of the chain". This could also be accomplished via the management >>plane. Perhaps a discussion of control plane and management plane >>could be added to section 2.2? From the balance of the draft, it seems >>that control plane is deferred for a future proposal? >>=20 > >PQ> Precisely. We expect there to be several possible control plane >options depending on environment and use case and this is something the >WG would be assigned to work through. >Ron> My suggestion is that the concept be explicitly supportable with a >control plane and without (i.e., solely via coordinated management of the >affected Network Elements). This may allow for earlier adoptions and >thereby market validation of the concepts. SK> By not requiring a specific model, NSH is generic to be implementable in either model. As Paul said, further work in the WG can add to this. > > >> 3.1 NSH Actions; 3. Update a Service Header - this section introduces >>the concept of a service index. From context, I assume that this >>indicates the previous node was considered to be index N out of M total >>nodes? Is it 0-based or 1-based? > >PQ> Correct, the index provides location with the path, and is >decremented. Service index=3D0 means the packet should be discarded. >Ron> I suggest that the text be enhanced to bring out the semantic >meaning more explicitly. > > >>=20 >> 3.3 NSH Usage; 2. Service Chaining - this section introduces OAM >>messages, but none are defined elsewhere in this document. Does this >>lend itself to a separate draft? If so, should this document >>acknowledge that it won't attempt to define OAM procedures? Perhaps >>it could acknowledge the possibility that the fields following the Base >>Header would be different for OAM messages? >>=20 > >PQ> Yes, the plan is to have a future OAM focused document. I'll update >the draft to reflect that. > > >> 3.3 NSH Usage; 3. Metadata Sharing -It is mentioned that semantics of >>the metadata require coordination via a control plane, but this seems >>somewhat contrary to the previous statement that a fixed size and format >>header is used for simplicity. > > >PQ> The size is fixed, but the semantics of the data itself needs to be >coordinated by the control plane. For example, if the network shared >context header is used to convey some form of classification, the value >carried in the context but be meaningful to all participating nodes. The >control plane assigns that meaning. >Ron> Suggest that the concept include the option whereby the management >plane assigns that meaning, too. > > >>=20 >> 3.4 NSH Proxy Nodes - it is mentioned that the proxy may deliver the >>packet to the service node via a local attachment circuit. I suggest >>that it should be specified that the service node is required to deliver >>the packet back to the proxy (i.e., "bump-in-the-wire" forwarding >>paradigm) via the same or different access circuit. Also, this >>mechanism could also be extended to non-NSH tunnels (i..e, IP-in-IP or >>GRE). >>=20 > >PQ> Good clarifications, and yes, the same of different circuits is >viable, as are other encaps. I'll update the doc. > > > > >> 4. Header Format; C bit - the text states that when C bit is set, there >>are one or more contexts present. But then states that the format is >>fixed to 4 32-bit contexts. > >PQ> This is a parsing optimization: there are always 4 contexts but if >non are used, then c=3D0 and they don't need to be checked. >Ron> This was more of an editorial comment. "When the C bit is set, all >four contexts are present" would more accurately reflect the intention, >if I have interpreted correctly. > > >>=20 >> 4. Header Format - the distinction of network/service and >>platform/shared is not clear to me. Perhaps some usage examples could >>be added to help clarify? >>=20 > >PQ> Very valid comment, the next revision will include examples. > > > >> 4. Header Format - for deployment topologies that encompass access >>networks (e.g., most), it is very common to wish to convey a subscriber >>identity (e.g., IMSI or MSISDN in a mobile network) distinct from the IP >>address used in the packet. This may facilitate application of local >>policy. It can also facilitate application of local policy in network >>functions chained after a NAT function. Perhaps an optional 128-bit >>generic subscriber identity can be considered in the NSH? On the >>other hand, I can see where keeping the NSH size minimal is extremely >>beneficial. One thought here would be a mechanism that could be >>employed by stateful network functions using flag bits to acknowledge >>that certain metadata has been received and is no longer necessary in >>subsequent packets of the fully qualified flow. With this approach, >>the 128-bit generic subscriber identity could be conveyed in the first N >>packets to the next node until a packet is received back from the next >>nod > e with the corresponding acknowledgement bit set. This approach could >be extended to other fields, including the 4 context fields already >defined. >>=20 >> =20 > >PQ> The header is flexible enough to accommodate such a case but I need >to understand the use case in more detail. In theory, a control plane >could indicate that the context headers are to be interpreted >differently. Again, I'd like to get a better sense of how you'd use this. >Ron> Given this split notion of classification-capable nodes and >non-classification-capable nodes, could the classification nodes add more >value than simply choosing which SF's need to be invoked. In >particular, certain SF's would be capable of providing differentiated >services for different groups of subscribers. Can the classifier >determine not only that some SF needs to be inserted but also which >locally significant set of policy should be applied? If so, how is this >conveyed from the classifier to the SF? Inband? Out of band? SK> Precisely! The classification-capable nodes can provide such metadata or ID to the service nodes to provide enhanced service delivery where the service node is acting on a policy selected by the classifier-capable node. This ID may in turn lead to fine-grained classification at the service-node based on any number attributes, including, flow, environment, network, ... How to carry large metadata seems like a perfectly valid discussion to have as part of the WG and should be input to the WG when it is formed. > > > >>=20 >> _______________________________________________ >> nsc mailing list >> nsc@ietf.org >> https://www.ietf.org/mailman/listinfo/nsc > >_______________________________________________ >nsc mailing list >nsc@ietf.org >https://www.ietf.org/mailman/listinfo/nsc